Cognitive Accessibility of Racial
Stereotypes, Beliefs, and Self-esteem
in Black and White College Students
A Thesis presented to
The School of Graduate Studies
Drake University
In Partial Fulfillment
of the Requirements for the Degree
Master of Arts
by
Deanna Carroll
March 1994
Cognitive Accessibility of Racial
Stereotypes, Beliefs, and Self-esteem
In Black and White College Students
by
Deanna Carroll
Approved by Committee:
Chairperson
-
- -
Dean of the School of Graduate Studies
I would like to thank my dear family and friends.
Without your
support and continued encouragement, 1 would have never made it through
this endeavor.
I am eternally grateful to Dr. Judith Allen, whose support
and patience prevailed through endless revision after revision.
I would
like to express my sincere appreciation to the members of my committee,
whose support was never ending. I also wish to thank other members of
the faculty and administration, who never stopped believing in me and
cheering for me. Thank you all from the bottom of my heart.
Cognitive Accessibility of Racial Stereotypes,
Beliefs and Self-esteem in Black and White College Students
An Abstract of a Thesis by
Deanna Carroll
March 1994
Advisor: Judith Allen
The problem.
The purpose of this study was to assess whether internalization of racial stereotypes
existed in black and white college students. It was hypothesized that cognitive accessibility of
racial stereotypes would occur and would be displayed by differential reaction times to
stereotypes following the racial categories of black and white. In addition, the cognitive
accessibility of racial stereotypes was hypothesized to be related to levels of self-esteem and
racial beliefs.
Procedure.
-
The sample was composed of 76 college students (39 Black 20 male, 19 female, and
37 white - 14 male, 23 female) who performed a semantic priming task (Dovidio, Evans, and
Tyler, 1986; Dovidio, Perdue & Gaertner, 1991) and completed questionnaires assessing selfesteem and racial beliefs.
Findings.
An internalization pattern of positive versus negative stereotypes emerged. Irrespective
of race, black and white subjects associated positive more than negative stereotypes with the
category of white and negative more than positive stereotypes with the category of me. There
were no correlations between levels of self-esteem and internalization. Overall all subjects had
an above average level of self-esteem, independent of whether they displayed any internalization
of stereotypes. There were correlations between levels of racism and the degree to which
subjects associated positive versus negative words to the two categories of people and the
category of self. Those subjects exhibiting higher levels of racism had slower reaction times to
positive stereotypes, when they followed the black and me prime.
Conclusions.
The findings of this study suggest that the internalization that existed was one of an
evaluative nature. Subjects internalized the association of positive and negative stereotypes,
regardless of levels of self-esteem. This was different from our prediction that subjects would
display internalization of black and white stereotypes.
Recommendations,
Re evaluate the stereotypes to assess how they are defined by a sample of black students.
It also may be more beneficial to examine the effects of internalization of positive versus
negative traits.
TABLE OF CONTENTS
INTRODUCTION
h/lETHoD
RESULTS
DISCUSSION
REFERENCE
APPENDICES
.................................................................................................
...................................................................................................
...................................................................................................
.................................................................................................
...................................................................................
...................................................................................................
A.
Coopersmith's Self-Esteem Inventory
B.
Modern Racism Scale
C.
Informed Consent
D.
Instructions to participants
................................... ....
.......................................
.......................................
.......................................
LIST OF TABLES
TABLE
I.
PAGE
Percentage of Error Responses ('No" responses to
Correct Prime-Adjective Pairs) to Positive and Negative
Adjectives as a function of Racial Category and Self
Primes.
2.
Percentage of Error Responses to White and Black
Stereotyped Adjectives as a function of Racial
Category and Self Primes.
3.
Percentage of Error Responses to Positive and Negative
White and Black Stereotyped Adjectives as a function of
Racial Category and Self Primes.
4.
Mean Reaction Times (in milliseconds) to Racial and Self
Primes.
5.
Mean Reaction Times (in milliseconds) to Positive and
Negative Evaluative Adjectives as a function of Racial
Category and Self Primes.
6.
46
Mean Reaction Times (in milliseconds) to Racial Category
and Self Primes as a function of Race of Subject and
Evaluative Nature of the Adjective.
47
Stereotypes, Beliefs and Self-esteem
1
Introduction
The concept of a stereotype has been included in the psychoIogical
literature for some time.
One of the earliest definitions was provided by
Lippmann, who suggested that stereotypes were distortions, caricatures,
and institutionalized misinformation.
In addition to this, Lippmann further
depicted stereotypes as pictures in our heads (Lippmann, 1922). Sherif and
Cantril (1947) and Secord (1959) suggested that stereotypes were a special
category of attitudes composed around and toward some given object or set
of objects, without any basis in experience
OF
knowledge.
LaViolette and
Silvert (1951) discussed stereotypes as emerging from social interaction
and therefore, as being social attitudes in the strictest sense.
They also
believed stereotypes to be composed of two main attributes, those of
persistence and rigidity;
persistence referring to their continuation across
generations or duration through time, and rigidity indicating that
stereotypic attitudes are extremely resistant to change despite conflicting
evidence (LaViolette & Silvert, 1951).
Vinacke (1949) defined stereotyping
as the tendency to attribute generalized and simplified characteristics to
groups of people in the form of verbal labels and to act, according to those
labels, in a certain way towards the individuals in the group.
Allport (1958)
described a stereotype as an exaggerated belief associated with a category.
Tajfel believed stereotyping to involve 'the attribution of general
I
psychological characteristics to large human groups' ( 1 969, p. 81-82).
While these investigators have displayed some important differences in the
way they have defined a stereotype or the process of stereotyping, they all
agree on most of the defining features: it is an overgeneralization, a
category concept that is learned, factually incorrect, rigid and persistent
(Ashmore & Del Boca, 1981). In sum, a stereotype may be defined as a set of
Stereotypes, Beliefs and Self-esteem
2
beliefs about the personal attributes of a group of people based solely on the
individual's membership in the group (Ashmore & Del Boca, 1981). Ashmore
and Del Boca further explain that knowledge of the stereotypes does not
mean belief in them.
Instead, stereotypes are regarded as a component ot an
overall body of knowledge an individual possesses surrounding certain
groups in the society. In this paper, I will focus on stereotypes ascribed to
blacks, the extent to which knowledge of these stereotypes is learned
and/or accepted by black individuals in our society, and consequences of that
internalization of stereotypes for self-esteem.
Ethnic attitudes and stereotypes are part of the social heritage of a
society (Ehrlich, 1973).
Ehrlich argues that stereotyping is the structuring
of the elements of belief statements about groups of people.
He further
explains that this structuring of elements of belief statements about groups
of people is inherent in the basic history ot society.
Ehrlich feels that
stereotypes are transported across generations as an element of the
accumulated knowledge of society, and 'that no individual can grow up in a
society without learning the stereotypes attributed to the major ethnic
groups.
Given this assumption that stereotypes are communicated over
generations as an element of the accumulated knowledge of society and that
any individual growing up in this society would inevitably acquire knowledge
of such stereotypes, theorists believe that internalization of these
stereotypes would occur even in members of stereotyped groups (Devine,
1989).
Internalization refers to the process of taking in information and
making it an integral part of one's attitudes or beliefs (The American
Heritage Dictionary, 1982).
It is theorized that black individuals, as well as
white individuals, acquire knowledge of stereotypes concerning blacks
merely from their existence in the society, and that these individuals may
make this stereotypic information an integral part of their attitudes or
Stereotypes, Beliefs and Self-esteem
3
beliefs regarding themselves.
Both empirical and theoretical work in the
developmental psychological literature generally suppori this assumption
that children will internalize cultural stereotypes about their own social
groups (Clark 8r Clark, 1947; Goodman, 1952).
Theories and Research on Internalization of Stereotypes
Proshansky (1966) concluded from empirical studies that ethnic
attitudes begin to form at the age of three or four, in both black and white
children.
During this period and as a part of the larger process of
establishing a sense of self, children develop an awareness of their own
ethnic identity.
The child becomes aware that ethnic distinctions are made
and that he or she as well as others are perceived as belonging to differing
social groups.
The child's racial awareness increases rapidly for the next
several years and is well established by the time the child enters first
grade.
Porter (1971), and more recently Katz (1976), agree that most
children become cognizant of race and ethnic attitudes during the nursery
school years.
Goodman (1952) studied 57 black and 46 white nursery school
children between the ages of three and five. This study employed a number
of observational procedures and play-interview techniques; for example,
using pictures of children and dolls to determine how a child would respond
to questions dealing with levels of racial awareness.
Goodman found that
85% of both the black and white children had some awareness of racial
characteristics, and that this racial awareness increased with age.
Clark
and Clark (1947) in their study of 253 black children between the ages of
three and seven also found racial awareness to occur early and increase with
age.
Other researchers have obtained findings consistent with those of
Goodman and the Clarks (i.e., Ammons, 1950; Horowitz & Horowitz, 1938;
Landreth & Johnson, 1953; Morland, 1958; Stevenson & Stevenson, 1960;
Stereotypes, Beliefs and Self-esteem
4
Stevenson & Stewart, 1958; Vaugh, 1964).
Issues which continue to be under debate in the literature include the
nature of the mechanisms which determine the content of racial attitudes,
and whether or not these attitudes include internalized stereotypes about
one's own social groups in *the same way that stereotypes about others'
social groups become internalized.
Tajfel and Turner (1 986) provided a very different explanation for the
internalization of stereotypes.
Tajfel and Turner developed Social Identity
Theory, in which they stated that attitudes are assigned to groups based
simply on an individual's membership or lack of membership in a group. The
"in-group" consists of people with whom the individual identifies himself or
herself, whereas the "out-group" consists of everyone else.
Their general
model suggests that people tend to attribute positive characteristics to the
group to which they belong and to the individuals within their own group,
while they attribute negative or less positive characteristics to everyone
else; the out-group and its members. Tajfel and Turner believed the
explanation for this ingroup enhancement-outgroup bias was that people
derive a large part of their self-identity from their social identity.
Social
identity, in large part, comes from the status of the social groups to which
the individual belongs.
It consists of those aspects involved in self-image
that derive from the social categories to which people perceive themselves
as belonging.
As a result, these social groups provide their members with a
sense of identity.
These identifications are relational and comparative.
This means that they define the individual as similar to or different from,
as 'better' or "worse" than members of other groups.
Individuals strive to
maintain or enhance their self-esteem, and therefore, they strive for a
positive identity.
Since this positive social identity is based on favorable
comparisons that can be made between the in-group and some relevant out-
Stereotypes, Beliefs end Self-esteem
5
groups, the in-group must be perceived as positively differentiated from the
relevant out-groups.
In turn, this leads to people attributing positive
characteristics to the group to which they belong and negative or less
positive characteristics to the relative out-groups.
Given 'that individuals
wish to achieve and maintain a positive self-identity, they will be
motivated to see their own social groups as better, and others' social groups
as worse.
This theory is not always descriptive of the judgments of minority or
subordinate group members.
Milner (1975) along with Tajfel and Turner
(1986) found that minority or subordinate group members frequently
derogate their own group and display positive attitudes towards the
dominant out-group.
In addition, subordinate groups often derogate
themselves in self-evaluations.
Since dominant group members derogate
their out-groups, which would be subordinate groups, these negative
evaluations are the dominating beliefs in the society.
The subordinate group
members internalize these beliefs, which leads to negative evaluations of
themselves rather than the normal in-group appreciation evaluations.
This
would suggest that everyone internalizes .the concepts put forth by the
dominant group, even when these concepts are negative evaluations of
groups to which one belongs (Tajfel & Turner, 1986)
Other theorists hypothesize that parental attitudes alone influence
racial attitudes in children.
In the case of prejudice, it is most commonly
believed that children are prejudiced because their parents are prejudiced,
and have subsequently transmitted these feelings and attitudes to their
children (Katz, 1976). While this claim would seem reasonable, evidence has
not s~rpportedits validity.
For example, Radke-Yarrow, -Trager, and Miller
(1952) discovered that kindergarten and first grade white children often
displayed negative reactions to blacks even when their parents possessed
Stereotypes, Beliefs and Self-esteem
6
more liberal attitudes.
Radke-Yarrow, Trager, and Miller interviewed 101
parents of first and second grade children in four public schools in
Philadelphia and compared the parental attitudes held towards differing
social groups to those attitudes held by their children.
They found that a
positive or neutral attitude towards a group held by the parents generally
led to a higher percentage of their children rating that group as positive or
neutral in their attitudes.
However, this was not the case when assessing
white children's attitudes toward black children.
In spite of positive
parental attitudes, white children exhibited negative attitudes toward black
children (Radke-Yarrow, Trager & Miller, 1952).
Additional recent evidence
of the seemingly weak relationship between parentaf racial attitudes and
the racial attitudes of their children was found by Branch and Newcombe
(1986).
They found that children of black activists showed slightly more
white preference than the children of non-activists.
Branch and Newcombe
reasoned that in homes in which parents have strong pro-black attitudes,
they may confront the reality of race as an issue in the society in a clearer
way than do parents with less strong beliefs.
In turn, this initially may be
disturbing to young children, who learn that the dominant culture has
negative attitudes toward their race.
Later, after these issues are
processed more fully, this same family environment may be a source of
strength.
Bird, Monachesi, and Burdick (1952), Frenkel-Brunswik and Havel
(1953), and Pushkin (1967) have likewise found there to be no relationship
between parental attitudes toward social groups and the attitudes of their
-young children. While the early attitudes of young children do not seem to be
related to those of their parents, over time these attitudes change and the
environment provided by parents is found to play a role in this change
(Branch & Newcombe, 1986).
The mechanism of reinforcement also has been suggested as a method
Stereotypes, Beliefs and Self-esteem
7
of racial attitude acquisition, but there is not much evidence to support this
idea (Katz, 1976). Proponents of this view say that peers and adults in the
child's environment are seen as positively rewarding the expression of
negative attitudes.
It is argued, however, dhat racial attitudes are
composed of complex responses not typically acquired under conditions of
consistent reinforcement, since the communication of these attitudes is
often ambivalent, and may lead to inconsistent reinforcement (Katz, 1976).
Katz provides an example of this reasoning, involving a white first-grader,
Jane, in an interaction with her mother. Jane asks her mother if she can
invite her friend Lotus, who is black, over for lunch. The mother says "yes,"
but uses a different tone of voice than that generally heard by the child.
Then Jane asks her mother if she can go to Lotus' house the next week;
however, the mother says "no," and tries to rationalize her answer to the
child.
"The message, then, that often gets transmitted to the child is a
confusing one with many overtones" (1976, p. 131).
Katz (1976) stated that the establishment of ethnic attitudes was
related to the development of a child's self-identity.
According to Katz,
children learn the groups to which they do and do not belong as a part of the
self-discovery process, and at this time positive and negative feelings are
associated with various groups.
Katz explained prejudice as being
attributable to an authoritarian parenting style.
This
model stated that
prejudice in children was generated by the environment provided by harsh
and rigid parents.
Porter (1971) stated that one of the most importand methods of
attitude transmission was the family.
She drew this conclusion based on
the idea that since children would not be born with social values and
attitudes, topics beyond their comprehension, they would have no alternative
but to internalize the values, norms, and behavioral patterns of parents or
Stereotypes, Beliefs end Self-esteem
8
others.
Porter (1971) said that as the individual would grow older, he or she
would learn not only how to group people into social categories, but also
what these classifications would mean in terms of social desirability.
In
addition, Demo and Hughes (1990) argue that socialization experiences,
particularly parental messages were important in shaping racial identity.
They further stated that the family context was generally felt to be the
most influential socialization setting for forming the child's emerging sense
of self, values and beliefs.
Their findings from a study using the National
Survey of Black Americans supported the idea of racial identity being shaped
by the content of parental socialization.
If internalization of cultural stereotypes is linked to the social
environment, then one would predict that as these environmental factors
change, the individual's degree of internalization of attitudes should also
change over time.
Even though the attitudes of the children's parents do not
seem to be the sole cause of the types of attitudes their children will
develop, the overall environment created by the parents may have an effect
on the child" attitudes.
Environmental factors for blacks have changed
the American
since the onset of the Civil Rights Movement. C~~rrently,
socio-cultural environment includes more positive images of blacks in the
society, and an overall concern for the social welfare of black individuals
(Rosenberg & Simmons, 1971).
Assuming that black children internalize
stereotypes as a consequence of learning from the culture, there should be a
corresponding pattern of change in the degree of internalization of black
stereotypes by black individuals.
Indeed, this pattern of change can be seen
with a comparison of studies demonstrating racial preference and
identification in black children prior to and after the Civil Rights Movement.
Stereotypes, Beliefs and Self-esteem
9
Research Prior to the Civil Riahts Movement. Clark and Clark ( I 939)
conducted the first study examining the degree of racial awareness and
attitudes of black nursery school children.
They modeled their study after
the Horowitz (1939) study examining children's self-awareness with
reference to specific social groupings.
Clark and Clark wanted to
investigate the degree to which black children would make selections
preferring a stimulus representing a black individual to one representing a
white person. They also wanted to measure the extent to which these
children would make correct identifications with the appropriate dolls
representing black individuals when asked to do so.
Clark and Clark showed
pictures consisting of various combinations of a white boy, black boy, lion,
dog, or clown, to 150 black children, three to five years old.
After the
instructions of 'Show me which one is you," for the boys and 'Show me
which one is
?" using the name of the girl's brother, boy cousin
or boy playmate for the girls,
they found that, while overall the children
made slightly more choices of the black child, this was not true at each age
level (1939, p. 594). When the responses were interpreted in separate age
levels, the findings showed that the youngest children chose the white child
in the picture more often (44% of the time) compared to the black child (41%
of the time) with 15% making irrelevant choices, such as one of the animals
or the clown.
The percentage of choices for the white child by black
children remained constant at 44% as the children go from three to four
years old; however, the children cease to identify themselves in terms of
the animals or the clown and consistently identify themselves in terms of
either the black or white children. This elimination of irrelevant choices of
the three year otds by ages four and five led to a trend in black children of
more choices of the black child with age. The black children chose to
identify with the black child in the picture 55% of the time at age 4, and
Stereotypes, Beliefs end Self-esteem
10
56% of the time by age five. As the children grew older from age 3 to age 5,
they increasingly identified with the black child.
Nevertheless, a
significant number of black children continued to identify with the white
doll: 45% at age 4, and 44% at age 5 (Clark & Clark, 1939).
Several years later, Clark and Clark individually interviewed black
children, ages three through seven, using a set of four dolls, two black and
two white.
They asked them the following questions:
" 1. Give me the dull that you want to play with.
2.
3.
4.
5.
6.
7.
8.
Give me the doll that is a nice doil.
Give me the doll that looks bad.
Give me the doll that is a nice cotor.
Give me the doll that looks like a white child.
Give me the doll that looks like a colored child.
Give me the doll that looks like a Negro child.
Give me the doll that looks like you" (1947, p. 602).
They felt the first four questions measured racial preference, while
questions five through seven measured racial self-identification.
Clark and
Clark found that black children preferred white dolls and rejected black
dolls when asked to choose which were nice, which looked bad, which they
would like to play with, and which were a nice color. Clark and Clark
believed this finding implied that the children perceived black not to be
beautiful. Hraba and Grant (1970) later interpreted these findings as being
consistent with the dominating racial tones of the times.
In 1958, Stevenson and Stewart completed a study using 125 white and
100 black subjects between the ages of three and seven years of age. They
studied racial awareness through doll assembly tasks, which required the
subject to assemble four two-dimensional dolls differing in color.
Second,
there were discrimination tests, which evaluated discrimination tasks
based on pictures displayed on cards. Third, there was a doll task, which
requested the subject to choose the doll that looked more like himself or
herself, and one with which he or she would rather play.
Finally, the fourth
Stereotypes, Beliefs and Self-esteem
11
task involved incomplete stories consisting of seven cards in color depicting
play situations, in which subjects were asked to answer questions
concerning the children depicted in the cards.
Stevenson and Stewart found
that the black subjects made a lower rate of own-race choices than did
white subjects in items involving the selection of a child as a playmate, as
looking most like the subject, as whom they wanted to go home with, and as
whom they wanted to attend their birthday party.
In addition, the black
subjects assigned negative roles to black children more frequently than the
white subjects assigned such roles to white children.
p
5
Goodman (1952) found in her study of 103 black and white children
that racial awareness was not only present at age three and four, but that
[
iK
25% of the children at age four were already showing strongly embedded
"ace-related
[
i
;
values.
For example, white children would never express a
desire to be like a black person, whereas black children exhibited a large
amount of denial and conflict regarding their evaluations of blackness.
r
Trager and Yarrow (1952) pointed out in their review of these earlier
studies that, while the tendency of white children to prefer white stimulus
figures in these experiments was not only congruent with the societal norms
but was congruent with positive self-concept; the black children's
preference for the white stimulus was congruent with the society's
dominant ideas, but did not allow for the development of a positive self-
[
t<
i
image.
In sum, these early studies indicated a white racial preference among
young black children and the tendency to attribute more positive
F
,
5
k7
t
$
ts
b
F
r
t
characteristics to whites than they did to blacks (Branch & Newcombe,
1986; Clark & Clark, 1939; Stevenson & Stewart, 1958; Trager & Yarrow,
1952).
As black children grew older, however, McAdoo (1971) found that
they developed an increasing black preference. These results suggested
I
Stereotypes, Beliefs and Self-esteem
12
that, even in the period prior to the civil tights, children's racial attitudes
were subject to change during the childhood years. These changes, McAdoo
suggested were related to social awareness, life experiences and the ability
to understand them.
4
The Civil Rights Movement was a time of open expression that it was
'okay" to be black. "In 1963, Isaacs presented some evidence that black
standards of beauty might be changing' (In Rosenberg & Simmons, 1971, p.
42). The Black Pride Movement had achieved considerable momentum by
1968 and was stressing the theme that 'black is beautiful' (Rosenberg &
Simmons, 1971). Grier and Cobbs are quoted as saying that "with the new
black power movements under way, all that we have just said [in the studies
of the 1940's and 1950Fs]may assume merely historical significance.
contorted efforts to be white, the shame of the black body,
...-
The
all may
vanish quickly" (1968, p. 54). As a consequence of this establishment of a
more positive image of blacks, black individuals could now achieve a more
positive belief system and a more positive self-concept.
With positive
beliefs, a black individual may still possess knowledge of stereotypes about
blacks, but should be less likely to believe them.
Research Durina and After the Civil Riahts Movement. Hraba and Grant
(1970) replicated the methodology of the Clark and Clark doll study of 1947
and found that a majority of black children ages three through eight
preferred a black doll, with this preference increasing with age.
They found
70% of the black children preferred the black doll as a playmate, 54% chose
the black doll as the "nice doll," and 69% selected the black doll as the doll
with the nice color.
Hraba and Grant also found that more of their subjects,
86% as compared to 72% of the subjects in Clark and Clark, made the correct
identification of which dolt looks like a Negro child.
Johnson (1966) found, from his experiments with 18 black youths in a
!.
Stereotypes, Beliefs and Self-esteem
13
Harlem freedom school, that these students rated black equal to white.
Gregor and McPherson (1966) found that Southern, urban black children 6 and
7 years old generally preferred a black doll.
Their results showed that 59%
of the black children chose the black doll as the one that looked the best;
60% selected the black doll as the one with the nice color; and 59%
identified the black doll as the "nice doll."
Results of a change in the image of blacks in society held by whites
has also been found.
Based upon the results of their empirical data looking
at whites' stereotypes of blacks and whites, Gaertner and McLaughlin (1983)
suggested that white college students no longer differentially associate or
ascribe negative characteristics to blacks, even though they may ascribe
more positive traits to whites.
In order to prevent responding indicative of
social desirability in this experiment, a lexical decision task was given to
subjects and reaction times were measured.
Subjects were unaware that
the experiment involved stereotyping when responding, and therefore, would
not have a reason to bias responses due to social desirability.
In addition,
Herrnstein (1990) further states that white attitudes toward blacks
indicate a pattern of change, as the principle of equal opportunity has been
generally accepted.
There have been various methodological problems in using pictures,
and especially dolls, to measure racial identity (Vaughan, 1986). One such
problem is found in the interpretation of the extent of misidentification
found in the Clarks' studies.
This issue of misidentification found by the
Clarks, where black children were reported as incorrectly identifying the
white doll as the one that looked like them, was further addressed by
Greenwald and Oppenheirn (1968). Greenwald and Oppenheim (1968) found
that with the inclusion of an intermediate (mulatto) skin color alternative,
there would be a reduction in the misidentification of the black children.
Stereotypes, Beliefs and Self-esteem
14
They performed an experiment modeled after Clark and Clark (1947), except
there were three dolls instead of two, a dark brown, mulatto, and white one.
They used 75 nursery school children, 39 black and 36 white. As
h~pothesized,the mulatto doll did play a role in reducing the percentage of
misidentification, only 13% of the black children compared to 39% found by
the Clarks' did not identify with the correct doll.
If we look at the light-
skinned black children, only 11% misidentified themselves as compared to
80% in the Clarks' study, and there was no misidentification found in the
dark-skinned black children, compared to 19% in the ClarksVesults.
Misidentification among white children was 44%.
However, Greenwald and
Oppenheim (1968) concluded that the mulatto doll could be an appropriate
choice for some white children as well.
With this taken into consideration,
they found the misidentification of white children to be only 19%.
Greenwald and Oppenheim (1968) concluded that the 19% misidentification
found in the white children was not significantly different from the 13%
misidentification found in the black children.
Their use of a white
comparison group and a wider range of response choices suggested that
there was nothing unusual about black children's misidentification as was
previously concluded by the Clarks (Greenwald & Oppenheim, 1968).
Even though misidentification may have been incorrectly interpreted,
there is still the issue of doll preference, and how black children responded
to questions of which doll looks nice in the 1930's as compared to how they
responded in more recent years. The trend would seem to indicate that, in
fact, responses in this area have changed. There are, however, studies that
would contradict this conclusion.
Gopaul-McNicol (1988) examined racial
identification and racial preference in 191 black preschool children in New
York and Trinidad.
The results reported a substantial majority of black
preschool children in both New York and Trinidad displayed a preference for
Stereotypes, Befiefs and Self-esteem
15
the white doll and identified with the white doll.
In addition, the children
more frequently chose the black doll as the one that "looks bad.'
It was also
noted that the children made disturbing remarks such as, 'l don't like being
black,' and 'I will be rich if I am like the white doll.'
This study could be
criticized again as making an incorrect interpretation of the
misidentification findings since an intermediate skin-colored doll was not
used.
In the case of doll preference, the same types of changes that were
occurring in other parts of the United States in terms of the Black Pride
Movement may have impacted these residents of New York and Trinidad to a
lesser degree.
Gopaul-McNicol (1988) found that while an area may be
predominantly black, this will not prevent white supremacy to prevail if
oppression is occurring.
This is exemplified by the fact that while
Trinidad's population is majorily black, the culture still possesses a
perception of white supremacy fostered in the media, the classroom and at
home. Consequently, many of these children are deprived of their share in
the pride that one would expect to come from being a merrrber of the
majority group in a population (Gopaul-McNicol, 1988).
In addition, Fine and Bowers (1984) also replicated the Clarks' (1947)
study with 58 black children, ages 4 to 6 in New Jersey. Their results were
also more similar to those of the studies during the period prior to the Civil
Rights Movement. They found 52% of the children preferred to play with the
black doll, 480A chose the black doll as the one that 'looks nice,' 52% chose
it as the one that ''looked bad,' 52% chose the black doll as the one that
'looks like a negro,' and 69% identified the doll as looking like themselves.
The authors concluded that their results could be an indication that black
children are growing more ambivalent about racial identity and are
experiencing the oppressive ideas that prevailed in the society before the
Civil Rights Movement (Fine & Bowers. 1984).
Stereotypes, Beliefs and Self-esteem
I6
I argue that perhaps there is another explanation as to why these
studies may not support the basic theme that the concept of blacks held by
blacks in the society has changed to be more positive. If change occurs at
different rates due to differential impact, then studies done in areas where
change may come slower will not produce the same results as studies done
elsewhere (Vaughan, 1986) For example, in Lincoln, Nebraska, where Hraba
and Grant (1970) performed their study, there had been a campaign to
promote Black Pride in the two years prior to their research. This could not
have been the case in New York, New Jersey, or Trinidad. The process of
social change in race relations in the United States may not impact the
entire nation at the same rate then nor now (Vaughan, 1986). In addition, by
~ gains of the Civil Rights movement may have
the mid to late 1 9 8 0 ' ~the
been fading in some communities.
Based upon the trends in past research, I believe that currently, while
blacks still have knowledge of the cultural stereotypes about blacks, it
seems that their beliefs in these stereotypes have decreased at least for a
period of time since the occurrence of the Civil Rights Movement.
In turn,
these studies since the Civil Rights Movement have depicted a greater
occurrence of the rejection of negative black stereotypes by black children
(Hraba & Grant, 1970; Johnson, 1966; Gregor & McPherson, 1966; but see
Gopaul-McNicol, 1988; and Fine & Bowers, 1984). While, in general, there
appears to be a greater rejection of stereotypes, recent contradictory
findings could suggest a regression in what seemed to be a positive change
in the attitudes of black individuals.
White knowledge of the culture's
stereotypes about blacks inevitably may be acquired by black individuals,
the degree to which individuals believe in these stereotypes may be
determined by other socialization processes. That is, although one may have
knowledge of a stereotype, his or her personal beliefs may or may not be
Stereotypes, Beliefs and Self-esteem
17
congruent with the stereotype (Devine, 1989).
If the individual is socialized
to possess a positive belief system regarding an ethnic group, he or she will
be less likely to employ in his or her belief systems the negative
characteristics of the stereotype, even though this individual may have
knowledge of such characteristics.
Ashmore and Del Boca (1981) also made
note that contemporary social scientists do not equate knowledge of
stereotypes with belief in them.
Instead, stereotypes are regarded as a
component of an overall body of knowledge an individual possesses in
regards to certain groups in the society. 'This body of knowledge may be
consistent with the beliefs of the individual, or it may be inconsistent,
which would lead the individual to dispute rather than believe in the
information contained in the stereotype (Ashmore & Del Boca, 1981).
Nevertheless, even though the individual may not believe in the stereotypic
information, this information may still be cognitively accessible to the
person, and therefore used in processing information about individuals in
that group when conscious inhibition is prevented.
Social Cognitive Models of Internalization of Stereotypes
In recent years theorists have focused on understanding how
knowledge, including stereotypes, is represented in memory.
Current
theories suggest that all knowledge acquired by individuals is contained
within complex knowledge structures, sometimes termed schemata (Fiske &
Taylor, 1984). Fiske and Taylor (1984) defined a schema as a cognitive
structure that represents organized knowledge about a given concept or type
of stimulus.
These organized knowledge structures contain information that
is congruent, as well as incongruent, with the overall content of the
knowledge structure.
For example, an individual's knowledge structure
Stereotypes, Beliefs and Self-esteem
18
concerning blacks may include knowledge of positive and negative
stereotypic information as well as attributes that may contradict the
StereOtYPes. They reasoned that people use these schemas to guide how they
encode, retrieve, and make inferences about raw information that they
receive from the environment.
People have a tendency to simplify reality by
interpreting specific instances in light of the schemas they possess, and to
draw conclusions about a particular situation, object or person from
information already existing in their memory.
Further, this organized prior
knowledge enables one to function in a social world that otherwise would be
perceived as too complex (Fiske & Taylor, 1984).
This grouping of information can be found with reference to social
categories (Ehrlich, 1973).
grouped and classified.
Ehrlich feels that all social objects can be
He states that this shared classification of social
objects represent both individual and societal responses to the management
of the complexity of the culture.
social categories.
nationality.
These shared classifications are termed
Examples of social categories would be race, religion or
These examples represent classifications of people based on
distinctive, easily recognized and easily articulated criteria shared by
members of the society.
Stereotypes are descriptive characteristics or
traits assigned to individuals based on their membership in a particular
social category (Ashmore & Del Boca, 1981). A schema about a particular
social category of people will include stereotypes, if any exist with regard
to the category in the culture.
This model would explain internalization as a
process in which an individual incorporates the Stereotypes of the culture.
in addition to their own trnique experiences with members of a given
category, into the corresponding social schema.
individuals will then
interpret the behavior of and react towards members of these social
categories based, in large part, on the information contained within their
1
I
Stereotypes, Beliefs and Self-esteem
19
schemas,
t h e s e schemas possessing stereotypical information are
described as conservative, but not necessarily rigid structures, so that
change, while difficult, is possible.
This understanding of knowledge structures offers an explanation as to
why differences in the internalization of negative stereotypes of blacks
occurred in some black children after the Civil Rights Movement. Over time,
attitudes regarding blacks in the society changed and began to include more
positive information contradictory to the negative stereotypes.
People
eventually began to incorporate additional characteristics, contradictory to
the stereotypes about blacks, as a part of their revised schemas.
With these
modifications, in some instances, negative evaluations could be moderated
(Hraba & Grant, 1970; Rosenberg & Sirnmons, 1971).
Negative evaluations are not always moderated, however, because the
processing of stereotypic information appears to be somewhat more complex
than described above. There is some evidence that once the content of our
social schemas is extremely well-learned, it becomes so embedded in our
knowledge structures that we may process or use this information
relatively automatically.
Bargh (1989) defines an automatic process to be
"an effortless sequence of processing that is developed from frequent
experiences within a particular stimulus domain, and triggered by the
presence of this stimulus without the necessity of conscious intent or
control" (page 2). Since automatic processing does not involve conscious
intent or control, the information used in this type of processing is
comprised of those data which are most accessible to the individual.
He
denfines more conscious or controlled processes as involving consciousty
processed thoughts, awareness of what information is being evaluated, and
control over one's responses to this evaluation.
Conscious processes are
flexible and easily adaptable to the particular features of the current
Stereotypes, Beliefs and Self-esteem
20
situation.
Devine (1989) reviewed evidence (Katz, 1976; Porter, 1971;
Proshansky, 1966) that stereotypes are well established in children's
memories before children develop the cognitive ability and flexibility to
question or evaluate the validity of the stereotypes. As a consequence, she
suggested that one's personal beliefs about the appropriateness of
stereotypic ascriptions are newer cognitive structures than one's
stereotypic knowledge sttructures.
Stereotypes are believed to have a longer
history of activation and therefore are more chronically accessible than are
personal beliefs.
In addition, because the stereotype has been frequently
activated in the past, it is a set of well-learned associations (Dovidio,
Evans, & Tyler, 1986).
These well-learned associations are automatically
activated and become highly accessible in the presence of a member of the
stereotyped group (Devine, 1989).
Devine further said that this
unintentional activation of the stereotypes is equally strong and equally
inescapable whether or not the person believes in the stereotype. The mere
presence of the target stimulus will, in a sense, trigger these chronically
accessible constructs (e.g., stereotypes), so that the person is more likely tto
use stereotypic information while processing information at that time.
An
individual who does not believe in a particular stereotype, and who deems it
an inappropriate basis for behavior or evaluation will experience a conflict
between the automatically activated stereotype and his or her own personal
beliefs.
Because the stereotypic associations have a longer history of
activation than the more newly acquired personal beliefs, rejection of the
stereotype requires intentional inhibition of the automatically activated
stereotype and activation of the newer personal belief structure.
inhibition and isritiation requires controlled processing.
This
Therefore, Devine
Stereotypes, Beliefs and Self-esteem
21
concludes that whereas stereotypes are automatically activated, activation
of personal beliefs requires conscious processing.
This need for conscious processing in order to reject the stereotypes
is a consequence of the individual's having to overcome a lifetime of
socialization experiences which have taught the stereotypes of the culture.
Therefore, if a stereotype is automatically activated in the presence of a
member of the stereotyped group, and an individual who does not believe in
the stereotype cannot or does not consciously monitor this activation, this
activated stereotypic information will influence subsequent processing of
information.
Devine suggested that this is the way in which knowledge of
stereotypes could influence responses even of those individuals who do not
endorse the stereotypes.
Further, Devine stated that in order for the
individual to begin rejecting stereotypes at an automatic level, he or she
must increase the accessibility of the cognitive structures that represent
the newer set of associations which do not utilize the negative stereotypes.
As a result, Devine theorized that if these new associations occur often and
over time, the information in the belief structures of the individual should
become as accessible as the stereotyped information.
In order to test these theories, Devine (1989) first performed a study
in which she looked at the extent to which high- versus low- prejudice
individuals differed in their knowledge of cultural stereotypes of blacks.
Devine had forty white introductoly psychology students participate in her
study.
She created the high- and low- prejudiced categories by
administering to all subjects McConahay's (1986) seven-item Modern Racism
Scale.
This scale has proven to be useful in identifying an individual's level
of racism.
Subjects were assigned to a high- or low- prejudice group based
on how they scored on the scale. Then she asked the students to list the
characteristics forming the stereotype of black individuals.
Stereotypes, Beliefs end Self-esteem
22
Devine found no difference between the high- and low- prejudice
subjects' description of the cultural stereotypes concerning blacks.
indicating equal knowledge of the stereotype by all individuals.
There were
no statistically significant differences in the components of the
stereotypes that were reported by the two groups, and judges could not
identify the high- prejudice subjects' listings from the low- prejudice
subjects' judging from the content of their lists.
Although Devine's findings
are based on white subjects, I believe that these findings would apply to any
individual in the society as suggested by Ehrlich (1973).
In Study 2, Devine (1989) examined both the effects of level of
prejudice (i.e., high- versus low- prejudice) and automatic stereotype
priming on subjectshvaluations of ambiguous stereotype-related behaviors.
These were performed by a race-unspecified target person under conditions
that precluded the possibility of controlled processes explaining the effects
that resulted.
Automatic stereotype priming is the automatic activation of
stereotypic information due to the presence of an activating stimulus or
prime. A prime is a word representing the stereotyped group.
Results
suggested that automatic stereotype activation was equally strong and
equally inescapable for high- and low-prejudiced subjects.
Devine found
that when the subjects' ability to consciously monitor stereotype activation
was impaired, both high- and low- prejudiced subjects produced stereotypecongruent or prejudiced-like responses.
As Bargh (1989) has suggested,
those traits which were strongly associated with a social category, in this
case blacks, were chronically accessible to the perceiver in the presence of
a target stimulus.
In Study 3 , Devine's (1989) results provided evidence that controlled
processes can inhibit the effects of automatic processing, or chronically
accessible constructs, when knowledge of the stereotypes is not congruent
Stereotypes, Beliefs and Self-esteem
23
with individuals' personal beliefs.
In this study, subjects were asked to list
their thoughts about the racial group of blacks under anonymous conditions.
it was found lhat under these conditions, high- and low- prejudiced subjects
wrote different thoughts about blacks.
For example, high-prejudiced
subjects gave responses such as "Blacks are free loaders," or "Blacks cause
problems," as opposed to such responses such as "Blacks and whites are
equal," or "Affirmative action will restore historical inequities," as given
by low-prejudiced subjects.
According to this model, although low-
prejudiced individuals have changed their beliefs concerning stereotyped
group members to be more egalitarian, the stereotypes have not been
eliminated from their memory structures.
Instead Devine demonstrated that
the stereotype remained a well-orgarrized, frequently activated knowledge
structure.
Applying these constructs of relatively automatic to relatively
controlled processing of information to the phenomenon of internalization,
it would seem that the degree of internalization of stereotypes can be
measured by at least two different cognitive factors: (1) the relative
cognitive accessibility of the relevant stereotypic traits to the individual,
and (2) the relevant belief systems of the individual.
Previously cited
definitions and studies of internalization (e.g., Clark & Clark, 1939, 1947;
Katz, 1976; Goodman, 1952; f roshansky, 1966) have essentially represented
internalization as being beliefs or attitudes about one's own social
categories, or identification with another, higher status social categoly.
In
establishing a more complete description and understanding of
internalization, the degree of cognitive accessibility of stereotypic
information must be taken into consideration, as well as conscious beliefs.
While some Endividuals may not consciously endorse the stereotypes of the
culture about their own stigmatized group(s), this stereotypic information
Stereotypes, Beliefs and Self-esteem
24
still may influence the process of self-evaluation, judgment, decisionmaking, and behavior, to the degree that this information is cognitively
accessible to them.
If individuals automatically assign stereotypes to
groups, this is indicative of a high degree of accessibility of those
stereotypic knowledge structures.
Responses made at a relatively more
automatic level would not allow for active, conscious inhibition or
evaluation of the stereotypic information, but instead, would result in
decisions based on the relatively accessible stereotypic information rather
than the person's more egalitarian belief system.
If, however, the person
does not consciously endorse these stereotypes, then controlled processing
would be more likely to result in thoughts and decisions which do not
indicate stereotypic judgements.
Therefore, complete internalization might
be described as both an integration of knowledge of relevant stereotypes
into knowledge structures which are readily accessible, and conscious
endorsement of this information.
Conversely, an individual in whom
stereotypic information is no longer internalized is one for whom this
information is no longer readily accessible, and is not consciously endorsed.
This individual still has knowledge of the stereotypes, but this knowledge
simply is no longer easily accessible, and is not believed. Various degrees
of partial internalization c o ~ ~ exist
ld
when individuals do not consciously
endorse the relevant stereotypes of the culture; but still access this
stereotypic information in some decisions, as a result of this information
being relatively accessible.
All individuals who grow up as a part of a stigmatized group are
believed to experience some degree of internalization along a continuum
from complete internalization and chronic accessibility of these
stereotypes to total rejection of stereotypic beliefs and only relative
accessibility of stereotypic knowledge structures (Atkinson, Morton, & Sue,
StereoQpes, Beliefs and Self-esteem
25
1989).
In order for individuals to move towards total rejection of
internalized stereotypes, where stereotypic information is not accessible as
well as not consciously endorsed, two difficult tasks must be undertaken.
First, not only must belief structures about one's own group be changed to
become positive, but the more difficult task of strengthening the cognitive
accessibility of stereotypic-incongruent information must occur as well.
These theories suggest that for a person's chronically accessible
stereotypic responses to change, the person needs to frequently activate and
think about his or her egalitarian personal beliefs.
The individual must
increase the frequency with which the personal belief structure is activated
when responding to members of the stereotyped group, so that the contents
of the personal belief structures become more chronically accessible than
the older stereotypic knowledge structures.
The more chronically
accessible the individual's personal belief structures become, the more
likely they are to provide a strong, rival response to these responses that
would normally be expected to follow from more automatic stereotypic
activation.
Before the contents of Lhe newer belief system can be activated
at a relatively more automatic level, the previously well-learned
association between stereotypic attitudes and blacks will have to be
weakened and the association of black individuals with the new counterstereotypic, positive beliefs will have to be made stronger.
In conclusion,
the belief change process requires intention, attention, and time.
The
individual must not only inhibit automatically activated stereotypic
information, but he or she rr~ustalso intentionally replace such an activation
with responses congruent with the new belief system (Devine, 1989).
Summary
Devine (1989) concluded that knowledge of negative stereotypes about
Stereotypes, Belieis end Self-esteem
26
blacks is both inevitable and well-learned, so as to be chronically
accessible at an automatic level by American white individuals, regardless
of whether those individuals are egalitarian or prejudiced towards blacks in
their personal beliefs.
Theories of internalization (Ehrlich, 1973) would
suggest that this process of learning negative stereotypes about blacks may
be just as inexorable for American black children as for white children,
Based upon my review of the recent trends in empirical findings on
internalization of stereotypes, I believe that even though the beliefs of
black individuals concerning blacks have changed to be more positive since
the Civil Rights movement, the cognitive accessibility of negative
stereotypic information for many black individuals has not been reduced in
comparison to the accessibility of contradictory, counter stereotypic
information.
This idea of chronic accessibility of stereotypic knowledge structures
as a part of the internalization process has not been evaluated in blacks.
Devine established a theoretical framework by which to evaluate such a
phenomenon in her experiments involving high- and low-prejudiced white
s ~ ~ b j e c t s . I will utilize this framework in the investigation of cognitive
accessibility of black and white stereotypes for black subjects.
Cognitive accessibility of stereotypes will be measured through an
analysis of reaction times of subjects in response to stereotypic traits
following the names of groups using a modification of the priming procedure
developed by Dovidio and colleagues (Dovidio, Evans, and Tyler, 1986; Dovidio,
Perdue & Gaertner, 1991). Dovidio et at. used a computer task which
required subjects to make 'yes' or "no' responses to noun-adjective pairs
based upon whether the adjective could ever be descriptive of the noun. The
nouns were black, white, or house and the adjectives were traits descriptive
of houses or traits descriptive of people that were identified as black or
Stereotypes, Beliefs and Self-esteem
27
white stereotypes.
They found their white subjects displayed faster
reaction times to white stereotypic traits and positive characteristics
when they followed the white prime than when they followed the black
prime.
In addition, subjects had faster reaction times to black stereotypic
traits and negative characteristics when they followed the black prime than
when they followed a white prime.
I will modify the Dovidio et. al.
procedure in a pattern modeled after Dovidio, Perdue, and Gaeriner (1991) by
reducing the amount of time subjects view the prime and by adding a
backward mask to make the task relatively more automatic.
F~trther,a
category prime of "me" will be added to access any differences in an
individuals' internalized knowledge of the stigmatized group to which they
belong, and their perceived self (as a member of such a group).
Hypothesis 1
The Doll studies and effects of Civil Rights Movement suggest that the
stronger the association of positive beliefs with blacks, the less accessible
negative attitudes or beliefs should be at both the automatic and controlled
levels of processing.
I agree that the degree to which negative attitudes
concerning blacks are chronically accessible by blacks will not be as great
as in the past before such positive changes as brought about by the Civil
Rights Movement.
not be as abundant.
And in turn, acceptance of negative stereotypes also will
However, I do hypothesize that black college students
will demonstrate some evidence of chronic accessibility of stereotypes and
negative attributes about blacks in their knowledge structures concerning
blacks.
As a result, black subjects will have faster reaction times to black
stereotypic and negative adjectives after the primes black and me than after
the prime white, and faster reaction times to white stereotypic adjectives
Stereotypes, Beliefs
end Self-esteem
28
after the prime white than after black or me.
Black subjects also will
respond faster to negative traits and black stereotypes as compared to
positive traits and white stereotypes, after the primes black and m e
The Implications of Self-esteem for Internalization of Negative Stereotypes
Differences in self-esteem levels of black individuals may be related
to internalization of negative stereotypes by blacks.
Self-concept is
described as a construct consisting of conscious and unconscious feelings of
the self (Samuels, 1977). "It encompasses all that a child brings to the
statement 'This is me,'" and includes an understanding of one's attitudes,
capabilities, and the feelings accompanying these self-perceptions (Phillips
& Zigler, 1980, p. 1 12). These feelings and thoughts are believed to be
constantly changing by new learning and experiences from the individual's
interaction with his or her environment (McDonald, 1980; Phillips & Zigler,
1980; and Samuels, 1977). The evaluative component of the self-concept is
self-esteem.
McDonald (1980) and Samuels (1977) defined self-esteem as
involving the acceptance of and respect for the self.
Porter (1971) and
Rosenberg and Simmons (1971) described self-esteem as one's basic sense
of personal worth and adequacy. Allport (1937, 1943, 1955) and Bloom
(1964) suggested that self-identity and self-esteem begin to form as a part
of racial awareness during the preschool years.
According to Allport, the
three-year old has achieved bodily sense, self-identity, and self-esteem
(Samuels, 1977).
Bloom (1964) suggested that the most rapid growth and
development of the personality occurs in the first five years of life, and
that the basic self is acquired within the family in these early years
(Samuels, 1977).
Although they start to form early, self-attitudes and self-
conceptions are not static but are constantly changing and developing in the
Stereotypes, Beliefs and Self-esteem
2Q
Process of building self-esteem (Felker, 1974).
Yeatts (1 968) and Beglis
and Sheikh (1974) found that the content of the self-concept changed with
age, which resulted in more abstract and symbolic identifications as the
child grew older.
The concept of self-esteem seems to be more complicated for ethnic
minorities than for dominant groups, because, in addition to the individual
achieving an identity of self, they must also achieve an identity as a member
of an ethnic minority group. This process of ethnic identity development has
been shown to have an effect on levels of self-esteem in minorities.
Porter
(1971) suggested that the devaluation of one's racial group, as was the case
in the past for blacks in the society, created feelings of inadequacy and
insecurity regarding the self and that this was still happening after the
black power movements.
Porter conducted a study testing his theory that
due to the low evaluation of blacks in American society, black children
would have less positive personal identities than do whites.
significant difference in the overall scores by race.
He found a
Whites had better
personal self-concepts than did blacks, with sixty-three percent of the
whites versus thirty-one percent of the blacks having high self-esteem
(Porter, 1971).
Clearly, some black children were learning cultural
evaluations of race that unfavorably affected their self-esteem regardless
of the Civil Rights movement (Porter, 1971).
Other research of the 1970's and 1980's (e.g., Clark, 1985; Rosenberg &
Simmons, 1972) supported a belief that the evaluation of blacks in the
society would affect the level of self-esteem of black individuals.
However,
this research showed a change in the trend of self-esteem levels of blacks.
it was shown that black Americans did not have lower self-esteem or more
negative self-concepts than white Americans, as was the case in the past
(Clark, 1985). Theorists argued that white America was no longer the
Stereotypes, Beliefs and Self-esteem
30
looking glass for black America and that black Americans were internalizing
positive attitudes from their immediate environment, such as their family,
and neighborhoods (Clark, 1985; Wyne, White 8 Coop, 1974). Beglis and
Sheikh (1974) suggested that the impact of the 'black nationalism"
movement, with the emphasis on 'black is beautiful," and the appearance of
more blacks on television and in advertising had a considerable effect on
this generation of black children.
They f o ~ ~ nthat
d the self-concepts of the
black children varied slightly in terms of the content, but not in terms of
the extent to which positive evaluations were made.
Beglis and Sheikh
believed that the self-concepts of black children were being affected by the
new social climate, and perhaps there was an increasing emphasis on racial
pride in their self-perceptions.
Empirical evidence has been supportive of
their claims.
For exampte, Yeatts (1968) conducted a study of 8,979 children from a
Florida public school system, of which approximately one-third were black,
with all socio-economic levels represented.
Yeatts evaluated the subjects'
responses to the Gordon's "How I See Myself" self-report scale.
The results
rejected the hypothesis that self-reported self-concept would vary by race.
lnstead they found that self-reported self-concept was not a unitary
concept, and it varied by sex and age.
Rosenberg and Simmons (1972) found that, while black subjects found
lighter skin color to be more attractive regardless of whether they
themselves had darker skin, this preference did not affect whether they
considered themselves physically attractive.
Sixty-three percent of the
black children and 65% of the white children said that they were either 'very
good-looking' or 'pretty good-looking.'
The subjects' evaluations of
themselves as attractive were related to self-esteem.
They found no
differences in the overall levels of self-esteem in blacks versus whites.
Stereotypes, Beliefs end Self-esteem
31
The'$ reasoned that the belief structures of Blacks had begun to change,
allowing them to enjoy greater levels of self-esteem.
Samuel$ and Griffore (1 979) conducted a study of 72 black, white, and
Mexican-American boys and girls and evaluated their self-esteem using the
Purdue Self-concept Scale for Preschool Children. They too found no
differences in self-esteem among black, white or Mexican-Americans.
Wyne, White, and Coop (1974) suggested that there was a shift from a
self characterized by a sense of "nobodiness" to a sell filled with
"somebodiness;" that is, a change occurred in blacks from evaluating the
self against white significant others, to an evaluation of self that is clearly
and proudly reflecting of black significant others.
They further stated that
this increase in an orientation towards a positive black image helped black
individuals overcome the negative stereotypic imagery ever present in
society.
Porter (1971) felt that lower levels of self-esteem in black children
restJted from the negative evaluations of blacks in the society.
The changes
in the image af blacks to a more positive one is posited to have resu!ted in
higher levels of self-esteem in blacks (Beglis & Sheikh, 1974; Rosenberg &
Simmons, 1972; Samuels & Griffore, 1979; Wyne et. al., 1974; Yeatts, 1968).
Since the evaluation of self would also involve the evaluation of blacks by
black individuals, this higher level of self-esteem may be related to the
individual's increased rejection of the negative Stereotypes concerning
himself or herself, and blacks in general, rather than the internafization of
them.
While Porter (1971) found self-esteem levels of blacks to be lower
than their white counterparts, Yeatts (1968) and others (e.g., Beglis &
Sheikh, 1974; Rosenberg & Simmons, 1972; Samuels & Griffore, 1979) found
self-esteem levels of blacks to be no different than those levels found in
Stereotypes, Befiefs and Self-esteem
32
whites.
Theorists who have critiqued the literature on racial preference and
self-esteem levels of minority ethnic groups say that since these studies
have been conflicting in their findings, they are not a sound basis from
which to draw conclusions (Banks, 1976). Banks suggested that there is a
need to discard such studies as the Clark experiments because their findings
did not show significant differences, and therefore, may be a result of
chance.
However, to disregard years of research and findings consistent
among numerous researchers would be a greater error.
Instead of ignoring
inconsistent findings, one can seek to find an explanation as to why
differences in self-esteem levels reported in research on minorities has
resulted after the Civil Rights movement.
One explanation of why self-esteem levels vary among members of
ethnic groups involves the individual's level or position in his or her ethnic
identity development process (Cross, 1978; White & Burke, 1987). This
approach, referred to as identity theory, considers ethnic identity to be a
portion of the self that contains shared meanings of what it is to be a
member of a given ethnic group (White & Burke, 1987). Marcia (1980) and
Cross (1978) each identify a stage model of ethnic identity development.
In
Marcia's theory, Erikson's model of ego identity development is
operationalized for ethnic groups.
Erikson makes the point that members of
an oppressed and exploited minority, such as blacks, may internalize the
negative views of the dominant society, and in turn, develop a negative
identity and self-hatred.
Both Marcia's and Cross's models share the idea
that an achieved idenlity is the result of an identity crisis, which involves a
period of search or exploration, leading to a clear commitment (Phinney,
1989). The models differ, however, in that Marcia's paradigm does not
suggest a necessary developmental progression, whereas Cross' model
describes a progressive pattern of change in the identity of blacks.
Cross
Stereotypes, Beliefs and Self-esteem
33
describes blacks as progressing from lower to more advanced stages of
black identity.
Marcia developed four identity statuses, based on the presence or
absence of one's exploration and commitment to one's racial identity.
The
individual is characterized as having one of four types of identities: a
diffuse identity, foreclosed status, moratorium or achieved identity.
A
person with a diffuse identity has neither engaged in exploration nor made a
commitment.
A commitment made without exploration, usually on the basis
of parental values, characterizes an individual in the foreclosed status.
A
person in the process of exploration without having made a commitment is
in moratorium.
Finatly, an individual who has made a firm commitment
following a period of exploration is one who has an achieved identity.
An
achieved identity is characterized by a clear, confident acceptance of
oneself as a member of a minority group, replacing the negative self-image.
As discussed by Parham and Helms (1985), Cross (1978) identified the
process of developing a black identity as a progression through four distinct
psychological stages.
The person is believed to evolve from a perception in
which they degrade themselves for being black to a perception in which they
become secure about themselves as black people. The four stages, ranging
from least secure to most secure with oneself, are pre-encounter,
encounter, immersion-emersion, and internalization.
The pre-encounter stage is characterized by the individual viewing and
perceiving the world as being non-black, anti-black, or the opposite of black.
Individuals think, act, and behave in ways that devalue their black identity,
and experience whiteness as an ideal. In the encounter stage, the individual
experiences a shocking personal or social event that motivates them to
challenge their present frame of reference and become receptive to a new
interpretation of identity.
The person starts to change from an antiblack
Stereotypes, Beliefs and Self-esteem
34
world-view.
Immersion-emersion is the third stage.
Here, individuals begin
developing a sense of 'Black Pride." Their level of black awareness is high,
but the degree of internalized, positive attitudes about being black is at a
minimum.
Individuals tend to slander white people while glorifying blacks.
In the final stage, which is internalization, individuals obtain a feeling of
inner security and feel more satisfied with being black.
This occun when
aspects of the immersion-emersion phase have been incorporated into an
individual's self-concept.
reference for being black.
The individual has internalized positive frames of
This would coincide with Marcia's achieved
identity stage. There is a commitment to being black, and a secure
understanding and acceptance of one's identity (Krate, Leventhal, &
Silverstein, 1974; Parham & Helms, 1985; Phinney, 1989).
These earlier models of racial identity have come under criticism as
not showing the complete concept of racial identity development (Atkinson,
Morton, & Sue, 1989). Some of the criticisms of these models involve
conclusions that they unintentionally blame the victim, they have become
obsolete since they depend on societal factors that may have changed, they
are erroneous in assuming that identity development follows a linear and
continuous course and they are incorrect in making us view the stages as
static, discrete entities rather than a dynamic and evolving process
(Atkinson, Morton, & Sue, 1989).
In addition, Smith (1989) criticizes earlier
models as being based on racial self-hatred and a view of racial conflict
that is relevant but perhaps somewhat outdated. Therefore, one needs to
look for theories of development identifying positive racial identity
development.
Despite the criticisms, these earlier models provided a basis
from which more sophisticated models could be developed (Atkinson, Morton,
& Sue, 1989).
Due to the failings of the initial models of racial identity
Stereotypes, Beliefs and Self-esteem
35
development, researchers have in recent years modified the models in order
to address these shortcomings. The more recent model we will look at is a
modification of Cross' (1978) model and is put forth by Atkinson, Morton,
and Sue (1989).
They identify a minority identity model, which they feel
eliminates the flaws of the earlier models.
The model views minority
attitudes and behaviors as a product of an identity development continuum,
rather than a linear progression as suggested by Cross (1978). They feel
that an individual may exist in any one of the five stages without following
any progression. For example, a person could move from stage five back to
any one of the lower stages. Their mode! identifies five stages of
development that any oppressed group of people may experience as they
struggle to understand themselves in terms of their own minority culture,
the dominant culture and the oppressive relationship between the two.
The
stages are as follows.
Stage one is the Conformity Stage.
Minority individuals in this stage
are identified by their unequivocal preference for dominant culture values
over those of their own culture. There is a high desire to "assimilate and
acculturate."
In this stage, the attitudes of minorities about themselves are
those of devaluation and depreciation.
Individuals have group-depreciating
attitudes towards members of the same minority group, discriminatory
attitudes towards members of other minority groups and group-appreciating
attitudes towards members of the dominant group.
Stage two is the Dissonance Stage. The movement into this stage is
often gradual; however, as with Cross' (1978) model, a monumental event
may propel the minority individual into this stage.
In this stage minority
individuals begin to experience a breakdown in their denial system.
In this
stage, individuals experience a conflict between self-depreciating and selfappreciating attitudes, a conflict between group-depreciating and group-
Stereotypes, Beliefs and Self-esteem
36
appreciating attitudes, a conflict between dominant-held views of minority
hierarchy and feelings of shared experience towards members of different
minority groups, and a conflict between group-appreciating and groupdepreciating attitudes toward members of the dominant group.
Stage three is the Resistance and Immersion Stage. The minority
individual completely endorses minority-held views and rejects the
dominant society and culture.
Desire to eliminate oppression of one's
minority group becomes the motivation of the individual's behavior.
In this
stage, the individual has self-appreciating attitudes, group-appreciating
attitudes towards members of one's minority group, conflicting feelings of
empathy for other minority experiences and feelings of culturocentrism, and
group-depreciating attitudes towards members of the dominant group.
Stage four is the Introspection Stage.
The individual experiences
discontent and discomfort with group views rigidly held in the Resistance
and Immersion Stage, and the person's attitudes change to notions of greater
individual autonomy.
In this stage, the person experiences concern for the
basis of self-appreciating attitudes, concern with the nature of group
appreciation, concern with ethnocentric basis for judging others, and
concern with the basis of dominant group depreciation.
Stage five is the Synergetic Articulation and Awareness Stage.
Minority individuals experience a sense of self-fulfillment with regard to
cultural identity.
Desire to eliminate all forms of oppression becomes
motivation for behavior.
Feelings of discomfort and conflicts from the
Introspection Stage have all been resolved. The individual has selfappreciating attitudes, group-appreciating attitudes for members of ihe
same minority group, as well as for members of different minorities, and
attitudes of selective appreciating towards members of the dominant group,
~ l t h o u g hfive stages are presented in the model, Atkinson et. at.
Stereotypes, Beliefs and Self-esteem
37
(1989) state that their model is more accurately pictured as a continuous
process, in which one stage blends with another and the boundaries between
stages are not clear.
From the presented models on racial identity
development, one can see that an individual's position in the process of
racial identity development will affect the types of racial attitudes
towards self and others that the person will possess.
Through studying the identity development models, it has been shown
that the stage of an individual is related to this person's level of selfesteem. Paul and Fischer (1979) found racial acceptance among black
adolescents to be related to a more positive self-concept.
Parham and
Helms (1985) found that the encounter and internalization attitudes of
Cross' (1978) model were positively related to self-esteem.
Phinney (1989)
found minority adolescents who had achieved ethnic identity to have higher
scores in self-evaluation.
White and Burke (1987) discovered a significant
and positive correlation between commitment and global self-esteem.
The
more black individuals were committed to a black ethnic identity, the higher
was their self-esteem. In addition, Phinney and Alipuria (1990) found in
college students that ethnic identity commitment resulting in a confident
sense of self as a member of an ethnic group was significantly related to
self-esteem across four ethnic groups; not just blacks.
It would appear that self-identification, a sense of belonging, and
pride in one's group are crucial aspects of ethnic identity that are present in
varying degrees (Phinney, 1990). The extent to which individuals have
obtained an achieved ethnic identity has been clearly shown to be related to
self-esteem.
These results may explain why certain studies have reported
low self-esteem in blacks, while others have reported positive self-esteem
levels.
Cross and Marcia state that an achieved identity results from the
individual encountering a startling event that motivates them to search and
Stereotypes, Beliefs and Self-esteem
38
explore new interpretations of identity and then to commit to this more
positive interpretation as one's own ethnic identity.
The Civil Rights
movement can be identified as one key startling event that triggered for
many black Americans the start of what previously may have been an
undeveloped ethnic identity. Atkinson et, al. (1989) argue that not all
minority individuals experience the entire range of stages of racial identity
development.
Furthermore, prior to the Civil Rights Movement, a time in
which the transition of many individuals through the racial identity
development process was accelerated, many people may have lived their
lives in the first stage.
Currently however, some minority individuals may
be born and raised in a family functioning at level five and never appear to
experience a level-one sense of identity. On the other hand, it is their
opinion 'that many minority individuals are raised by parents functioning at
level five, but in coming to grips with their own identity, these children
often move from level five to one of the lower levels, and work through the
process for themselves.
From this I reason that more minority individuals
may be able to function at level five due to the positive impact of the Civil
Rights Movement (Atkinson, Morton, & Sue, 1989). During this period, many
blacks found role models in politics, sports, the media and a decrease in the
derogation of blacks (Paul 8 Fischer, 1979). Houston (1984) states that as a
consequence of the Civil Rights movement, black communities began to show
a markedly enhanced focus on unity, self-determination, power and racial
pride.
Blacks have been characterized as moving from a self-perception in
which they degraded themselves for being black, to one where they have
pride in being black (Parham & Helms, 1985).
Based upon these theories on ethnic identity development, researchers
should not assume that all blacks have either high or low levels of selfesteem or possess similar racial attitudes (Parham & Helms, 1985).
Blacks'
Stereotypes, Beliefs and Self-esteem
39
varying levels of self-esteem can be related to the extent to which they
have achieved ethnic identity. The stage in which a black individual exists
is believed to be affected by this person's socio-cultural environment.
While the Civil Rights movement may not have been felt in the same degree
by all individuals, as a result of its occurrence, more blacks seem to have
achieved ethnic identity.
Nevertheless, all blacks may not be at this point of
ethnic identity achievement, which would explain differences in selfesteem levels among black individuals in the different studies.
The present
research will not attempt to identify or predict levels of ethnic identity
development in subjects, but will focus on whether the degree of
internalization of racial stereotypes, defined in two ways, by relative
cognitive accessibility and degree of racist beliefs, will be related to
differences in levels of self-esteem in blacks.
Hypothesis 2
Individuals high in self-esteem are found to have a higher achievement
of ethnic identity (Parham & Helms, 1985; Paul & Fischer, 1979; Phinney,
1989; Phinney & Alipuria, 1990; White & Burke, 1987). This achievement of
ethnic identity is characterized by a secure understanding and acceptance of
being black, as well as internalized positive attitudes regarding being black
(Parham & Helms, 1985). Therefore, this higher level of ethnic identity and
higher self-esteem about oneself and one's ethnic group should be related to
rejection of negative stereotypes about blacks and acceptance of positive
traits by blacks.
In the first hypothesis, I predicted that black college
students would respond in a way indicative of relative accessibility of
stereotypes and negative attributes about blacks in their knowledge
structures concerning blacks.
As my second hypothesis, I further predict that high self-esteem will
Stereotypes, Beliefs and Self-esteem
40
be inversely related to the degree to which negative and/or black
stereotypes are chronically accessible to black subjects.
Specifically,
black subjects with higher levels of self-esteem are predicted to
demonstrate significantly slower reaction times to black stereotypes and
negative traits after the primes black and me, than will blacks with lower
self-esteem.
In addition, it is anticipated that these subjects will not
respond faster to negative, as compared to positive trait adjectives, nor
black, as compared to white stereotyped adjectives, after the prime black or
me.
If so, these responses will indicate that negative stereotypic
information for the categories of blacks and self is less cognitively
accessible for blacks higher in self-esteem than for those blacks lower in
self-esteem.
Additional secondary research questions were as follows.
Would
greater accessibility of positive and/or white stereotypes in white subjects
be related to higher self-esteem in whites? This is being done in order to
provide a comparison group to use in evaluating the results found in black
students.
Second, racial beliefs will be measured
in order to garner evidence
that while blacks may still have cultural stereotypes about blacks
cognitively accessible to them, they no longer demonstrate racist beliefs.
A
third research question will be addressed by evaluating black subjects'
cognitive representations of white stereotypes.
It is expected that black
subjects will show the same pattern of accessibility for the white
stereotypes as the white subjects, by displaying faster reaction times to
white stereotyped target adjectives, as compared to black stereotyped
adjectives, following the while prime.
White subjects will be used as a comparison group, in order to
determine if the patterns found in black subjects are different or similar to
Stereotypes, Beliefs and Self-esteem
41
those that will occur in white subjects at this time.
A fourth research question will be to determine the extent to which
black and white subjects consciously endorse the negative stereotypes
regarding blacks in the culture, in order to see if there is a relationship
between racial beliefs, accessibility of racial stereotypes, and self-esteem.
Method
Subjects
Originally 100 undergraduate college students at Drake University
volunteered to participate in this research.
However, due to insufficient
computer data and incomplete surveys, 24 subjects were eliminated from
the analysis (9 black subjects
-
subjects, and 12 white subjects
5 female and 4 males, 3 non-black minority
-
6 female and 6 male),
the majority being
eliminated due to incomplete surveys. Of the remaining 76, there were 39
black (20 male, 19 female)
and 37 white (14 male, 23 female) subjects.
Some students obtained extra credit points towards their Introductory
Psychology classes, while other students volunteered out of interest.
Materials
Coonersmith Self-Esteem Scale.
In order to assess self-esteem, the
Coopersmith Self-Esteem Adult Form was used. The Coopersmith SelfEsteem Inventories are widely used, with construct validity and reliability
well documented (Peterson & Austin, 1985). In a study of 103 college
students, Bedian et. al. (1977) reported Kuder-Richardson reliability
estimates (KR20's) of .74 for males and .71 for females. They also found
test-retest reliability estimates to be .80 for males and .82 for females.
Shaver and Robinson (1971) found adult form inter-item correlations to be
Stereotypes, Beliefs and Self-esteem
42
low for college students-
For 453 students, the average correlation was .13.
Convergent validity figures were found by Shaver and Robinson (1973). They
found correlations of -59 and .60 between the adult form and the Rosenberg
scale for college students (Coopersmith, 1990).
Semantic Priming Task. A priming task was run on Macintosh Plus
computers.
The task was a modified version of the Dovidio et. al. (1986)
procedure for demonstrating cognitive representations of racial stereotypes.
In this procedure, subjects were told that they would see a series of sets of
two words; a noun followed by an adjective. Their task was to decide if the
second word could ever be true of the first, and to answer as quickly as
possible by pressing a "yes" or "no" key.
The pairs of words consisted of a category prime and a test adjective.
There were four category primes: black, white, self, and house.
Sixteen test
adjectives were used; 8 words that describe a house and 8 that describe
people. To choose the 8 words that describe people, Dovidio and Gaertner
(1982) conducted a preliminary study which utilized an adjective checklist
procedure developed by Katz & Braly (1933), and Karlins, Coffman, and
Walters (1969). The Dovidio and Gaertner study asked 50 white males and
50 white females to identify out of 84 adjectives, 5 words they felt to be
descriptive of a typical white American and 5 words they felt were
descriptive of a typical black American.
From this, Dovidio et. al. (1986)
chose 4 words they felt to be most identified as being associated with the
white American, and 4 words felt to be associated most with black
Americans.
These 8 words were then identified as representing white
stereotypes and black stereotypes, respectively.
of
the 4 stereotypic words
for each group, 2 were considered positive evaluations and 2 were
considered negative evaluations.
8 words.
In the present research I used these same
This resulted in 4 categories of person-descriptive trait
Stereotypes, Beliefs end Self-esteem
43
adjectives: positive white Stereotypes (ambitious, practical), positive black
StereotY Pes (musical, sensitive), negative white stereotypes (stubborn,
conventional), and negative black stereotypes (imitative. lazy).
Modern Racism Scale. McConahay (1986) provided the following data
on the validity of the Modern Racism Scale. The validity of the Modern
Racism Scale was tested by evaluating how the items on the scale related to
voter preferences in the 1969 and 1973 Los Angeles mayorial contest
between Tom Bradley, the black challenger, and Sam Yorty, the white
incumbent.
The Modern Racism Scale correlated with voter preferences in
both contests.
Those whites scoring high on the scale were more likely than
low scorers to vote for the white candidate.
The correlations were ,392 in a
sample of white Claremont, California voters in 1969, ,365 in a sample of
white Los Angeles voters in 1969, and .338 in a sample of white Los Angeles
voters in 1973, all of which were statistically significant (McConahay,
1986).
The scale also correlated with strength of opposition to busing in
Louisville, Kentucky.
From surveys done during the conflict there in 1976
and 1977, correlations were 5 1 1 in 1976, and ,391 in 1977, these
correlations were statistically significant in both years (McConahay, 1982).
The Modern Racism Scale correlated ,383 with antiblack feeling as measured
by the Feeling Thermometer in Louisville (McConahay, 1982), and in Yale and
Duke University student samples, the correlations with the Feeling
Thermometer has averaged .441 for the past 16 years (McConahay, 1986).
Procedure
To begin, subjects were asked to read and sign an informed consent.
Subjects were told that the study was dealing with category assignments,
and that they would be asked to make some judgments at a computer.
Stereotypes, Beliefs and Self-esteem
44
Subjects first were allowed to familiarize themselves with the procedure
by performing a set of Practice trials.
Subjects then began the priming task.
On each trial, subjects saw One of four primes (black, white, me or house)
appear on the screen for 67 msec. The disappearance of the prime was
followed immediately by a mask; 'PPPPPn (for person prime) or 'HHHHH"
(for house prime) for 250 msec in order to make subjects less consciously
aware of the content of the words.
Following an interval of 250 msec the
target adjective was displayed for 250 msec (Dovidio, Perdue, Gaertner,
1991). All words appeared in the center of the screen.
Each person category prime (black, white, and me) was paired with
each of the 16 test adjectives twice, which resulted in 96 trials using
person categories. The house prime was paired 4 times with each test
adjective, resulting in 64 trials using house. The house prime was paired
more times than each of the person category primes in order to provide a
more equitable number of occurrences for "yes" and "no" trials.
pairings resulted in 160 randomly ordered trials.
These
For half of the subjects,
the "yes" key was on the right and the "non key was on the left; for the
other half of the subjects, the assignment of keys was reversed, There was
a brief intermission after the first 80 trials.
For half of the subjects, the
Coopersmith Self-Esteem Measure was given during this intermission.
For
the other half of subjects, the Coopersmith was not given at this time, but
instead was administered at the end with other instruments.
This variation
was done in order to test for an effect of order of the Coopersmith and the
priming task.
The intermission was included so that subjects would not get
too tired by completing all 160 trials at one time, which could result in
more random responding.
~t the conclusion of the priming task, subjects completed McConahay's
(1986)Modern Racism Scale in order to measure racial beliefs. In addition,
Stereotypes, Beliefs and Self-esteem
45
the subjects completed the Higgins Selves Questionnaire and the Rosenberg
Self-Esteem Questionnaire as a part of another study.
Upon completion of
these questionnaires, subjects were thoroughly debriefed about the
experiment and dismissed.
Results
Errors in Reaction Times
The percentage of 'no' answers in response to person-descriptive test
words that followed a white, black, or m e prime (for which 'yes' was the
appropriate answer) was 24%.
A multivariate analysis of variance
(MANOVA) performed on the error scores revealed a main effect of Prime,
F(1,76)=
-
29.96, ,Q<. 000 1, with subjects responding at a significantly higher
error rate to the black prime than the white prime, 1(76)=3.87, g<.001.
Subjects also responded with a sigr~ificantlyhigher error rate to the b l a c k
prime when compared to the m e prime, 1(76)=2.52, g<.025.
There was a significant Prime by Evaluation interaction, _F(1,76)=.5.8 1,
p<.005.
Subjects responded with a higher error rate to negative words
following the black prime versus the m e prime, 1(76)=3.10, ge.00 1, as well
as with the black prime versus the white prime, l(76)=7.15,
g<.00 1.
S ~ ~ b j e c talso
s
responded with greater error to negative words when they
followed the me prime as compared to the white prime, l(76)=4.06,
p<.001.
With regards to positive words, subjects responded with greater errors to
positive words following the black prime than the white prime, 1(76)=4.73,
~ 6 . 0 0 1 ,and the black prime compared to the m e prime, g76)=4.63, ~ c . 0 1.
0
...................................
Insert Table 1 about here
- - - - - - - - - - - * - - - - - - - - - - - - - - - - - - - - - - -
There was a significant Prime by Stereotype interaction, E(1,76)=
14.47, g<.0001.
Subjects responded with higher error rates to black
Stereotypes, Beliefs and Self-esteem
46
stereotypes following the white prime, 1(76)=2.28,
gc.05, and me prime,
-t(76)=2.83, p<.01, than they did to the white stereotypes. In comparison,
they responded with lower error rates to black stereotypes when they
followed the black prime. 1(76)=2.03, p<.05.
In addition. subjects responded
with more errors to the black stereotypes when they followed the black
versus white prime, 6(76)=2.99,
1(76)=2.08,
pc.05.
gc. 005, and the me versus white prime,
Likewise, subjects responded with more errors to the
white stereotypes following the black prime compared to the me prime,
1(76)=5.78, ge.00 1, and the black prime as opposed to the white prime,
1(76)=7.3, p . 0 0 1 .
----------------------------------Insert Table 2 about here
----------.----------------------Finally, there was a significant three way Prime by Evaluation by
Stereotype interaction, E(1,76)= 10.16, p<.000 1. Subjects made more errors
in response to negative black stereotypes following the me prime,
1(76)=4.38, e<.OOl, and positive black stereotypes when they followed the
white prime, f(76)=3.67,
of subject.
p<. 00 1. There were no interactions including race
Due to the systematic patterns of error found in the error data,
these data were removed and the correct answers to person-descriptive
words after person primes were analyzed separately.
This separate analysis
eliminated three of the original 76 subjects due to the amount of their
incorrect responses.
- - _ - * - - - - _ . - - - - - - - - - - - - - - - - - - - - - - - -
Insert Table 3 about here
-___------------------------------Correct Reaction Times
Post Hoc Scheffe, Stevens and Newman-Keuls Tests were performed on
the data initially, however, these tests were not effective in indicating
Stereotypes, Beliefs and Self-esteem
47
significant interactions of the data.
A 2 (Sex of Subject) by 2 (Presentation Order) by 3 (Prime) by 2
(Stereotypic Trait Type) by 2 (Evaluative Trait Type) by 2 (Yes Key) by 2
(Race of Subject) multivariate analysis of variance was performed on 73
subjects' correct response times to person-descriptive words after person
primes.
There were no main effects or interactions associated with sex of
subject, order of presentation of the Coopersmith Self-Esteem Inventory and
the priming task, or 'Mu versus "Z' as the yes key; therefore, these variables
were dropped from subsequent analyses.
There was a signiiicant main effect of Prime, E(1, 73)=15.54, g c . 0 0 5
(see Table 1). Overall, subjects responded faster to the black prime than to
the white prime, i(71)=3.2
(M" = 529.05 vs 706.8).
------------------------------.
Insert Table 4 about here
- - - - - - - - - - - - - - * - - - - - - - - - - - - - - - - - - - -
There was a significant main effect of Stereotype that emerged in the
initial analyses, E(1,73)=13.22,
g<.OOl;
however, subsequent post hoc tests
found no significant differences in reaction times to black versus white
stereotypes.
A significant Prime by Evaluative Trait interaction emerged, E(7, 73)=
8.15, ~ < . 0 0 1(see Table 2 ) . For both black and white subjects, positive traits
were responded to more quickly than negative traits following the white
prime, f(71)= 2.41, pc ,025 (M's = 653.42 vs 760.17), while negative words
were responded to more quickly than positive words following the me prime,
1(71)=2,22, p<.O5 (M's = 583.17 vs 681.43).
Overall, positive traits were
responded to more quickly following the black prime than after both the
white prime, 1(71)=2.$6, pc.01 (M's = 526.63
vS 653.421, and the m e prime
1(77)=3,496, g< .007 (M's = 526.63 vs 681.43) and negative traits were
responded to more quickly after both the black prime versus the white prime
Stereotypes, Beliefs and Self-esteem
48
f(71)= 5.17,pc.00 1 (M's = 531.47 vs 760.17) and the me prime versus the
white prime, f(71)=3.997,PC ,001 (M's = 583.17 vs 760.17).
However, there
were no differences in responding to negative traits following the black
versus me primes.
...................................
Insert Table 5 about here
_----_----------------------------The Prime by Evaluation interaction was qualified by a significant
three-way Race by Prime by Evaluation interaction E(1,73)=4.02,
pc.022 (see
Table 3).
Black Subjects.
Overall, black subjects responded more slowly to
negative words following the white prime than to positive words, 1 (36)=
2.279, pc .05(M's = 816.34 vs 676.5 ms), and more quickly to the negative
words as compared to positive words following the me prime, 1(36)= 3.19,
pc.005 (M's = 559.22 vs 755.26 ms). There were no significant differences in
their response times to positive versus negative words following the black
prime.
Black subjects responded more quickly to positive words following
the black prime versus the white prime, 1(36)= 2.13, gc.05 (M's = 545.74 vs
676.5 ms), and more slowly to positive words foIlowing the me prime than
the black prime, 2(36)= 3.414,gc.005 (M's = 755.26 vs 545.74 ms). There
were no significant differences in response times to positive words
following white versus me primes.
_
-
-
-
-
I
-
-
-
-
_
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Insert Table 6 about here
...................................
For responses to negative words, black subjects responded much more
slowly to negative words when they followed the white prime as compared
to the black prime, f(36)= 5.04, pc .001 (MIS = 816.34 vs 506.94 ms), and the
me prime 1(36)= 4.19,QC ,001 (M's = 816.34 vs 559.22 ms). There were no
significant differences in response times to negative words following the
Stereotypes, Beliefs end Self-esteem
49
black prime versus the m e prime.
However, there was a significant
difference between black and white subjects in reaction times to positive
words following the m e prime, 1(142)= 2.457, pc.025 (M's = 755.26 vs
601.26).
Overall, black subjects responded much more slowly to positive
words following m e than did white subjects.
White Sub-iects. For white subjects, there were no significant
differences in response times to positive versus negative words following
any of the three primes. There were no significant differences in responses
to positive words following the black prime versus the white prime, the
black prime versus the me prime, nor the white prime versus the m e prime.
However, white subjects did respond significantly more slowly to negative
words following the white prime as compared to negative words following
the black prime, 1(33)= 2.206, gc.05 (M's = 558.1 f vs 699.17 ms). There were
no differences in responses to negative words following the prime of black
versus m e , or white versus me.
Reaction Times and the Self-Esteem Inventory
The mean score on the Coopersmith Self-esteem measure was 76.82
for black subjects and 75.03 for white subjects, on a scale of 1 to 100, with
higher scores signifying higher self-esteem. The difference in these mean
scores was not statistically significant.
There were no significant
b
correlations found between self-esteem scores and differences in reaction
times for black or white subjects, although for white subjects, the
correlation between self-esteem scores and the mean difference in reaction
times of negative versus positive words following the me prime was
marginally significant, 1(37)=-,2507,pc.067. The greater the difference in
reaction times to positive adjectives minus the reaction times to negative
adjectives following the m e prime, the lower the self-esteem scores for
Stereotypes, Beliefs end Self-esteem
50
white subjects.
Reaction Times and the Modem Racism Scale
There was a significant difference in the mean score on the Modern
Racism Scale (MRS) for black versus white students, N76)=23.94 1, p<. 000 1 .
The mean score on the MRS was 2.38 for black subjects indicating a lower
level of racism, whereas for white subjects, the mean score was 5.95,
indicating a low although significantly higher level of racism than that of
the black subjects.
In order to assess the relationship between cognitive
representations of racial stereotypes and beliefs about racism, correlations
between reaction times to positive and negative stereotypes (black and
white) after the primes black, white, and me with scores on the MRS were
computed.
For black subjects there was one significant correlation between
scores on the MRS and mean reaction times. In response to the white prime
followed by positive white stereotypes, slower reaction times (which are
larger values) were associated with lower scores on the MRS (which
indicates less racism), ~(1,39)=-,2879,gc.038. 'There were no other
significant correlations for black subjects, but there were a few that were
marginally significant. The correlation between MRS scores and mean
reaction times to the black prime followed by negative black stereotypes
was -r(1,39)=-.2289, g<.08, indicating that the slower the reaction times to
negative stereotypes following the black prime, the smaller the score on the
MRS. t h e slower the subjects' responses to black stereotypes in general, the
smaller their scores on the MRS, _r(1,39)= -.2524, g<.061. Slower responses
to white stereotypes in response to the white prime resulted in lower
scores on the MRS, [(1,39)=-,2287, &!<. 08 1.
While there was orrly one significant correlation of response times
with MRS scores found for black subjects, there were several significant
Stereotypes, Beliefs and Self-esteem
51
correlations of scores on the MRS with reaction times for white subjects.
The results are summarized as follows:
Lower scores on the Modem Racism Scale were significantly
correlated with slower response times to:
positive black stereotypes following the black prime, 1(1,37)=--3492, pc.0 17,
positive black stereotypes following the white prime, ~(1,37)=-.3740, pc.0 1 1,
negative black stereotypes following the white prime, ~;(1,37)=-.3238, pc.025,
positive black stereotypes following the me prime, ~(1,37)=-.3970, pe.007,
positive white stereotypes following the me prime, 1(1,37)=-.3288,
pe.027,
negative white stereotypes following the me prime, ~(1,37)=-.3016, pc.035,
positive words following the bfaek prime, ~(1,37)=-.3017, p<. 035,
positive words following the white prime, 1(1,37)=.3475, pc.018,
positive words following the me prime, 1(1,37)=.3784,
pc.01,
negative words following the me prime, ~(1,37)=.3075, pc. 032,
black stereotypes following the black prime, ~(1,37)=-.3394, pc.02,
white stereotypes following the me prime, 1(1,37)=-.3479, p<,017, and
black stereotypes following the white prime, ~(1,37)=-.3877, p<.OOQ.
Overall, slower reaction times were significantly correlated with
lower scores on the Modern Racism scale for white subjects.
Discussion
The results showed that, overall, both black and white subjects
responded faster to all words following the black prime as compared to the
white and me primes, and slightly faster to the me prime in comparison to
the white prime. Consequently, conclusions are not easily drawn on the
reaction times between primes; for example, adjectives after black versus
white, or black versus me. However, we can evaluate responses to the
different adjectives following the individual primes.
In response to the
white prime, both black and white subjects responded more slowly to
negative words than they did to positive words.
These results indicate that
subjects associated positive words more than negative words with the
category of white.
This finding corresponds to Dovidio et. al.'s (1986)
Stereotypes, Beliefs end Self-esteem
52
findings, where subjects had faster reaction times to positive versus
negative traits following the white prime.
In addition, both black and white
subjects responded more quickly to negative words versus positive words
when they followed the me prime.
In this two way interaction, subjects
associated negative words more than they did the positive adjectives with
the me prime.
Finally, subjects did not associate either positive or negative
traits more with the black prime. These results for the black prime do not
agree with the findings of Dovidio et. al. They found their subjects
responded faster to negative words than they did to the positive words when
fotlowing the black prime.
Dovidio et. al. concluded that that their subjects
associated negative words more with the category of black than the positive
words.
These results, however, are qualified by race of the subject.
Black Subjects
Although black subjects' response times to positive as compared to
negative stereotypes were not significantly different following the black
prime, they were different for the white and me primes. After the white
prime, black subjects responded more slowly to negative as compared to
positive stereotypes.
Responses to positive as compared to negative
stereotypes were significantly slower following the m e prime.
This pattern
of reaction times suggests that black subjects associated positive words
more with 'the category white than they did negative words.
However, they
associated negative words with the category me more than they did the
positive words.
In addition, black subjects responded more slowly than did
the white subjects to positive words following the me prime. These
findings may indicate that black subjects associated positive words with
their idea of self less than the white subjects.
If one interprets "me" as
representing 'me as a black individual" for black subjects and 'me as a
Stereotypes, Betiefs and Self-esteem
53
white individual" for white subjects, the results can be compared to those
found by Dovidio st. al. (1986). in which white subjects responded to
positive adjectives more quickly following the white prime than when they
fol~owedthe black prime, and to negative words faster when they followed
the black prime than when they followed the white prime. In the present
study, black subjects' responses, both to the white prime and the me prime
as it applies to "me as a black individual," are similar to the pattern of
responses to positive and negative adjectives of white subjects in Dovidio
et. al.'s, after the primes of white and black. Dovidio, Perdue, Gaertner, and
Ellyson (1991) in their study of automatic activation of the "we" versus
Yhey" categories found that ingroup- and outgroup- designating words, such
as "we," "tey,"
and 'me," as in our study, can automatically and without
conscious awareness, activate differentially evaluative constructs in one's
memory.
However, in their study, 'the in-group is evaluated positively and
the out-group is derogated. The opposite has occurred with our black
subjects.
Black students showed evidence for derogation of themselves
the in-group, and positive evaluation of whites
- the out-group.
-
This pattern
of reaction times indicates evidence for internalization of negative
concepts associated with blacks or the self in our black college student
sample.
Because black children have grown up in a cultural context in which,
at a very early age they learn the same stereotypes as white subjects, one
can see how the results for black subjects in the current study resemble
those found by Dovidio et. al. (1986) in their white subjects.
While the pattern of reaction times to evaluative constructs following
the white and me primes for black subjects is consistent with a hypothesis
of internali+ation of negative constructs, the data do not totally support my
first hypothesis, that internalization of black Stereotypes would be
demonstrated in black subjects.
This would have led to differential
Stereotypes. Beliefs and Self-esteem
54
responding with regard to stereotyped adjectives, which did not occur.
There were no differences in responses to black versus white stereotypes to
the primes.
Instead, the internalization that was demonstrated involved the
association of positive traits with whites, and negative traits with the self,
for black subjects.
White Subjects
For white subjects, an interesting pattern of response times emerged.
There were no significant differences in responses to positive versus
negative words following any of the primes. They did, however, respond
significantly slower to negative words when they followed the white prime
versus when they followed the black prime, indicating that these subjects
more quickly associated negative words with the category of "black' than
with the category of 'white."
However, this difference in the average.
response time to negative words following the black versus white prime is
not a valid indicator of differences in construct accessibility in light of the
faster responses overalt to the black versus white and me primes, Because
of this main effect, overall reaction times were fastest in response to
stereotypes following the black prime, regardless of whether they were
positive or negative.
The fact that, overall, responses were very fast to all words following
the black prime could indicate that the presence of a black experimenter
(who conducted all experimental sessions) made the construct of black very
salient for both black and white subjects. The presence of a black
experimenter might have sewed as a prime to facilitate all subjects'
responses to the prime black. In addition, the presence of a black
experimenter may have caused subjects to be more cautious of racially
stereotyped responses. However, a subsequent study conducted by Recker
Stereotypes, Beliefs end Self-esteem
55
and Allen (1991) tested this effect of race of experimenter and did not find
a significant effect.
In comparison to the results of Devine (1989; Study 2), the results of
this experiment were only partially consistent.
She found that when her
white subjects' ability to consciously monitor stereotype activation was
precluded, both the high- and low- prejudice subjects produced stereotypecongruent responses.
In this study, the black and white subjects overall
showed a low level of racism (as indicated by the results of my subjects'
Modern Racism Scale scores). If I compare both the black and white subjects
to her low-prejudice category, I conclude that neither the black nor the
white subjects associated stereotyped words more with one category of
persons than the other. This would not indicate stereotyped responding
regardless of prejudice levels as found by Devine. However, while both black
and white subjects' levels of racism were low, the white subjects did
display a significantly higher level of racism than the black subjects.
Along
with this, 1 did find differential responding to positive versus negative
words, in my subjects.
Black subjects (lowest prejudice) associated
negative words more with the black and me primes and positive words more
with the white prime, whereas white subjects (higher prejudice) associated
negative words with the black prime and positive words more with white
and me.
Without considering racially stereotyped content, but instead using
positive and negative evaluative traits, one can see how these results are
partially consistent with those found in Devine's high and low prejudice
subjects.
t
i
r
Our black and white subjects displayed differential responding
indicative of internalization of evaluative constructs once the ability to
consciously monitor responses was minimized.
Stereotypes, Beliefs and Self-esteem
56
Error Data
The fast rate of presentation of stimuli may have resulted in higher
error rates than those found in Dovidio et. al. (1986). The significant
patterns in the error data that resulted also suggest that perhaps the
stereotypes used in this study are not sufficient to describe those held by
black subjects.
Overall, subjects responded with a higher error rate to
words following the black prime than they did to the white or me prime.
These stereotypes were prejudged by white subjects to be stereotypic of
blacks or whites,
These terms may not be seen or known by black subjects
as representing stereotypes.
While this might not be the case for our white
subjects, I believe that perhaps stereotypes defined by Dovidio et. al.'s
(1986) white subjects are not representative of those held by white
students in the current sampled population.
Self-Esteem Inventory
Analyses of the correlations between reaction times and scores on the
Self-Esteem lnventory did not s ~ ~ p p o my
r t second hypothesis, that selfesteem scores would be correlated with differences in reaction times, and
that internalization of stereotyped or evaluative traits might be linked to
self-esteem.
Overall, self-esteem scores were high for both black and
white subjects regardless of whether they displayed differential responding
to the primes. There are several explanations as to why this might be the
case.
For black students, self-esteem and racial attitudes could be two
separate and distinct concepts, in that the individual's idea of self might not
involve racial stereotypes.
Perhaps black parents are not teaching black
children that, 'you are black and because of this you should feel positive
Stereotypes, Beliefs end Self-esteem
57
about yourself," but instead, may be teaching their children to feel good
about themselves independent of their racial identity.
This would suggest
that the development of a positive self-concept would not involve thoughts
of one's racial identity but instead would involve thoughts of one's identity
as an individual.
In contrast, the literature reviewing racial identity
development as related to self-esteem has concluded self-esteem is
correlated with higher levels of achieved racial identity.
I would argue that
perhaps an individual cannot achieve racial identity according to the Social
Cognitive Model unless they first feel positive about themselves as an
individual.
This would explain why higher levels of self-esteem would be
related to higher levels of ethnic identity without self-esteem being a
function of one's racial concepts about the self. When comparing my
reaction time data to the pattern in self-esteem scores, black subjects
*,
responded more quickly to negative words following both the black and m e
prime than they did to positive words following these primes.
This finding
suggests that while their reaction times indicated a negative sense of self,
their self-esteem scores showed a positive sense of self.
As discussed
earlier, if individuals analyze the me prime as representing "me as a black
individual," and "me" as representing "me as an individual" on the selfesteem task, these two instruments may not tap into the same constructs.
On the other hand, one also could argue that the self-esteem instrument used
was a self repolt instrument. Therefore, subjects could have represented
themselves in a more positive light, while the reaction time data is less
open to impression management and may not represent beliefs about the
self, but rather 'knowledge' about the self in memory.
In addition, because
not all subjects were volunteers from an Introductory Psychology class, this
may have resulted in a more select group of high achievers, displayed by
their willingness to participate without the incentive of extra credit.
Since
Stereotypes, Beliefs and Self-esteem
58
the sample population was drawn from a private university, these conditions
could have resulted in a minimal or truncated range of differences in selfesteem among all subjects participating.
Modern Racism Scale
Results from the Modem Racism Scale supported the idea that blacks
do not believe in or consciously endorse the racial stereotypes of the culture
concerning blacks.
Whereas Devine recently found that some white subjects
did not endorse these racial stereotypes, mean scores of my white subjects
indicated that, on average, they believe in and consciously endorse the black
racial stereotypes of the culture at a level significantly higher than the
black subjects; however, not at a level indicative of strong racial beliefs.
Subjects displayed differential belief systems on the MRS, showing lower or
very low prejudice by black subjects and higher prejudice by the white
subjects.
These scores do not correspond with the indications found in
overall response times.
Devine (1989) argued that, regardless of beliefs, all
subjects wilt display stereotyped responding at an automatic level.
data only partially supported this assertion.
These
Black subjects did not respond
differentially to the racial nature of stereotypes in either their reaction
times or beliefs.
However, they did respond differentially to the evaluative
nature of stereotypes.
Subjects responded differently to positive versus
negative words following the racial primes.
Even though the white subjects
scored at a higher level of prejudice on the MRS, their responses to racial
stereotypes after the black versus white primes also were not indicative of
internalization of stereotypes utilized in this study.
I would argue that
while these subjects possessed some stereotyped beliefs about the outgroup the stereotypes used in this study were not indicative or
representative of these beliefs.
Thereby resulting in no differential
Stereotypes, Beliefs and Self-esteem
59
responding.
Dovidio et. al. (1991) would suggest that this in-group
facilitation is characteristic of aversive racism.
Aversive racism is
defined as a less direct form of racism whereby majority individuals do not
necessarily derogate the minority;
however, they view the majority as
being more positive than the minority.
Although my results partially replicated those found by Dovidio et. al.
(1986) with regard to differential responding to positive versus negative
words after the "white" and 'me" primes, I did not replicate the effect of
stereotype as was found in their study.
This failure to find stereotypes to
be facilitated by black and white primes supports the review of Doll studies
and racial identity.
I argued that from the trends in these studies, it would
appear that black individuals have changed not only their belief systems but
at least partially changed the way in which they internalize the stereotyped
information of the culture. And this change was felt to be towards an end of
elinrinating these constructs from one's concept of blacks as a group in the
society (Branch & Newcombe, 1986; Clark & Clark, 1939; Gaertner &
McLaughlin, 1966; Goodman, 1952; Gregor & McPherson, 1966; Herrnstein,
1990; Hraba & Grant, 1970; Johnson, 1966; McAdoo, 1971; Rosenberg &
Simmons, 1971; Stevenson & Stewart, 1958; Trager & Yarrow, 1952). The
internalization that has been demonstrated in my black subjects is that they
associate negative more than positive concepts with themselves.
It would
appear that black individuals are being socialized to develop more positive
constructs about the category of blacks as a whole but not necessarily about
themselves as a black individual. I suggest that these different methods of
socialization may not develop the same constructs surrounding a black
individual as it does blacks as a category.
This idea of blacks being capable of possessing a positive image of
self in a society where stereotypes prevail was verbalized by the author
Stereotypes, Beliefs and Self-esteem
60
Zora Neale-Hurston examined as she writes:
'I am not tragically colored . . . . I do not belong to the
sobbing school of negrohood who hold that nature has
somehow given them a low-down dirty deal and whose
feelings are all hurt about it. . . I do not weep at the
world - Iam too busy sharpening my oyster knife."
.
The society seems to have been effective in working on weakening the
effects of racial stereotypes.
However, stereotypes are but one component
of the overall concept of black individuals.
This concept also involves
whether to associate negative versus positive traits to this group of
individuals.
Our results indicate that while the stereotypes are not
pewasive in the concept of blacks, negative aspects still remain.
In this
area there is much that needs to be done in improving the overall concept of
i
I
blacks, and how they perceive themselves as a member of this group.
Another argument to explain the absence of stereotyped responding
would be that blacks may not 'stereotype' themselves, as they see the
diversity that is present in black individuals, which whites may fail to see
(Linville & Jones, 1980). According to Linville and Jones (1980), this
rationale for explaining why our black and white subjects did not display
stereotyping of themselves has to do with the existence of a more complex
schema concerning in-groups versus out-groups.
They define cognitive
schemas as prior knowledge structures that an individual develops through
experience (Linville & Jones, 1980). They propose that people have a more
complex schema regarding their own groups than other groups because they
perceive a large collection of diverse experiences involving persons in their
own group, whereas this is not the case for Out-groups. With a complex
schema, a large number of characteristics or dimensions will be used in
defining a stimulus.
Stereotypes, Beliefs and Self-esteem
61
In addition, another explanation would be that blacks may have
internalized different stereotypes than those measured in whites by Dovidio
et. ale (1986)- For the individuals in this study, perhaps all of the negative
adjectives (imitative, lazy, conventional, and stubborn) are a part of the
constructs they have associated with the category of black, but the two
positive adjectives (sensitive and musical) are not.
This would explain why
black subjects might not respond to the black stereotypes in the manner
hypothesized.
The black subjects also rrligfit show a pattern of differential
responding to negative versus positive adjectives as was the case in our
subjects. This is not to assume that black subjects do not have any positive
adjectives as a par! of their stereotypes but instead it is possible that the
stereotypes they possess do not include the particular positive adjectives
used in this study.
An implication for future research would be in the area of race of
experimenter.
The presence of a black experimenter in the current research
may have served as a prime, facilitating subjects' responses. Future
research should look for any effects of having an experimenter of a
particular race.
In addition, future research looking at internalization in black
students should redefine the stereotypes that are being measured to be ones
defined by a sample of black subjects, as opposed to the ones developed by
Qovidio et al.'s sample of white students. This should improve the
recognition of the words as stereotypes by black subjects and therefore
generate a better indication of the extent of internalization that may exist
in regards to stereotypes.
Along with this, one could also look at whether
internalization of stereotyped constructs varies for white versus black
individuals.
That is, whether black and white subjects encounter differing
stereotypes, and thereby develop constructs that are similar in evaluative
Stereotypes, Beliefs and Self-esteem
62
nature, but not in the stereotyped content.
Finally, future research may wish to consider the effects of
internalization of positive versus negative characteristics.
This would look
at whether students associate negative characteristics versus positive
characteristics more with a particular group, and whether this is consistent
for black and white students.
Tabfe 1
Percentage of Error Responses ('No" responses to Correct PrimeAdjective Pairs) to Positive and Negative Adjectives as a function
of Racial Category and Self Primes.
Primes
Black
White
Me
Positive
.3 14
.I81
.I84
Negative
,347
146
.260
AdieGtives
Total Number of Presentations = 160
Table 2
Percentage of Error Responses to White and Black Stereotyped
Adjectives as a function of Racial Category and Self Primes.
Primes
Stereotvped
pIdiectivie~
Black
White
Me
Black
.2975
,2005
,2680
White
,3635
1265
1760
Total Number of Presentations = 160
___________________---_--_----------------------------
Table 3
Percentage of Error Responses to Positive and Negative White and
Black Stereotyped Adjectives as a function of Racial Category and
Self Primes.
Prrmes
Black
ptdiective~
Black
Mean
SD
Positive
White
Me
.289
,311
White
Mean
SD
.339
.307
Black
Mean
SD
.306
.296
Mean
SD
,388
.304
Negative
White
Overall
,330
Total Number of Presentations = 160
___________________------------------------------~----
Overall
Table 4
Mean Reaction Times (in milliseconds) to Racial and Self Primes.
Primes
Blacb
Wbite
529.05
706.80
!.de
632.30
Table 5
Mean Reaction Times (in milliseconds) to Positive and Negative
Evaluative Adjectives as a function of Racial Category and Self
Primes.
White
Evaluative
Traits
Positive
526.63
653.42
681.43
Negative
531.47
760.17
583.1 7
Table 6
Mean Reaction Times (in milliseconds) to Racial Category and Self
Primes as a function of Race of Subject and Evaluative Mature of
the Adjective.
Race of Subiect
Black Students
Evaluative
Traits
White Students
Positive
Negative
Positive
Black
545.74
506.94
505.88
558.11
White
676.50
81 6.34
628.33
699.17
Primes
Stereotypes, Beliefs, and SeW-esteem
63
References
Allport, G. (1958). The nature of preiudice. Garden City :
Doubleday & Company, Inc.
The American Heritaae Dictionary. (1982). Boston : Houghton Mifflin
Company.
Arnrnons, R. (1950). Reactions in a projective doll-play interview
of white males two to six years of age to differences In
skin color and facial features.
76, 323-341.
I
Ashmore, R. and Del Boca, F. (1981). Conceptual approaches to
stereotypes and stereotyping. In D. L. Hamilton (Ed.) Cognitive
(pp. 1-35).
Hillsdale : Lawrence, Erlbaum Associates, Publishers.
Atkinson, D., Morton, G., and Sue, D. (1989). A minority identity
development model. counselins American Minorities: A
Crass-Cultural Perspective (3rd ed.). Dubuque : W.C. Brown.
Banks, W. (1976). White preference En blacks : a paradigm in
search of a phenomenon. P ~ v c h ~ l o ~ i Bulletin,
cal
83,
1179-1 186.
Bargh, J. (1989). Automatic and conscious processing of social
information. In J.S. Uleman, Unintended Thourn. New York :
Guilford Press.
Barnes, E. (1972). The black community as the source of positive self
concept for black children : A theoretical perspective. In
R. Jones (Ed.),Black Psvcholoc&!.
New York : Harper and Row.
Bedian, A. and Zmud, R. (1977). Some evidence relating to
convergent validity of Form B of Coopersmith's self-esteem
inventory. 5vcholoaical Re~orts,40, 725-726.
Beglis, J., and Sheikh, A. (1974). Development of the self-concept in
black and white children. Journal of Negro Educatioq, 43,
104-1 10.
Stereotypes, BelWs, and Self-esteem
64
Bird, C., Monachesi, E., and Burdick, H. (1952). Infiltration and the
hite and negro parents and children.
, 47, 688-699.
Bloom, B. (1964).
New York : John Wiley and Sons, lnc.
Bloom, H. (1986).
New York: Chelsea House Publishers.
Branch, C. and Newcombe, N. (1986). Racial attitude development
among young black children as a function of parental
attitudes : A longitudinal and cross-sectional study. Ghild
develop men^, 57, 712-721.
Clark, M.L. (1983). Social stereotypes and self-concept in black and
white college students.
125(6), 753-760.
Clark, K.B. and Clark, M.K, (1939). The development of consciousness
of self and the emergence of racial identification in negro
preschool children. The Journal of Social Psvchology, 10,
591 -599.
Clark, K.B. and Clark, M.K. (1947). Racial identification and
preference in negro children. In T. Newcombe , E. Hartley,
and E. Maccoby (Eds.),
New York : Holt.
Coopersmith, S. (1981 ).
San Francisco : W. H. Freeman.
Coopersmith, S. (1981). Self-Esteem Inventories. Palo Alto :
Consulting Psychologists Press.
Coopersmith, S. (1990). Self-Esteem Inventories.
Psychologists Press Inc.: Palo, Alto.
Consulting
Demo, D. and Hughes. M. (1990). Socialization and racial identity
among black americans. Social Psvchologv Quarterly, 53(4),
364-374.
Devine, P. (1989). Stereotypes and prejudice : Their automatic and
controlled components. The Journal of Perscnalitv and Socm
Psvcholoav, 56, 5-18,
Stereotypes, Beliefs, and Self-esteem
65
Dovidio, J., Evans, N.. and Tyler, R. (1984). Racial stereotypes : The
contents of their cognitive representations.
E
i
h
p, 22, 22-37. ~
Dovidio, J., Perdue, C.. Gaertner, S. and Ellyson, S. (1991). 'We' and
'They.'
The role of automatic activation in intergroup bias
Ehrlich, H. (1973).
John Wiley and Sons, Inc.
.
New York :
Felker, D. (1 974). Buildina Positive Self-Conce~ts. Minneapolis :
Burgess Publishing Company.
Fine, M. and Bowers, C. (1984). Racial self-identification: The effects
d
of social history and gender. Journal of A ~ ~ l i eSocial
Psvcholoay, 14(2), 136-146.
Fiske, S. and Taylor, S. (1984). Social Coanition. New York : Random
House.
Frenkel-Brunswik, E. and Havel, J. (1953). Prejudice in the
interviews of children : 1. Attitudes toward minority groups.
Journal of Genetic Ps~choloav,82, 91-136.
Gaertner, S, and Dovidio, J. (1982). The aversive form of racism, In
J. Dovidio & S. Gaertner (Eds.), Prejudice. Discrimination. and
Racism : Theory and Research. New York/Orlando : Academic
Press.
Gaertner, S. and McLaughlin, J. (1983). Racial stereotypes :
Associations and ascriptions of positive and negative
, 46, 23-30.
characteristics.
Goodman, M. (1952).
MA : Addison-Wesley.
.
Reading,
Gopaul-McNicol, S. (1988). Racial identification and racial
preference of black preschool children in New York and
Trinidad. The Journal of Black Psvcholoav, 14(2), 65-68.
Greenwald, H. and Oppenheim, D. (1968). Reported magnitude
of self-misidentification among negro children - artifact?
, 8(1), 49-52.
Stereotypes, Bellefs, and Se#-esteem
66
Gregor, A. and McPherson, A. (1966). Racial attitudes among white
and negro children in a deep-south standard metropolitan
, 68, 95-106.
area.
Grier, W. and Cobbs, P. (1968).
New York : Basic Books.
Herrnstein, R. (1990). Still an American dilemma.
Interest, 98, 3-17.
The Public
Houston, L. (1984). Black consciousness and self-esteem. Journal
, 11, 1-7.
Hraba, J. and Grant, G. (1970). Black is beautiful : A reexamination
of racial preference and identification.
and Social Psychology, 16, 398-402.
Horowitz, R. (1939).
school children.
Racial aspects of self-identification in nursery
, 7, 91-99.
Horowitz, E. and Horowitz, R. (1938). Development of social
attitudes in children. Sociometrv, 2, 301-338.
Johnson, D. (1966). Racial attitudes of negro freedom school
participants and negro and white civil rights participants.
Social Forces, 45, 266-272.
Johnson, El., Redfield, D., Miller, R. and Simpson, R. (1983). The
Coopersmith Self-Esteem Inventory : A construct validation
, 43,
study.
907-9 13.
Karlins, M., Coffman, T., and Walters, G. (1969). On the fading of
social stereotypes: Studies in three generations of college
, 13,
students.
1-16.
Katz, D., and Braly, K. (1933). Racial stereotypes in one hundred
coilege students.
28, 280-290.
Katz, p. (1976). The acquisition of racial attitudes in children. In
. New York :
P.A. Katz (Ed.),
Pergarnon Press.
Stereotypes. Beliefs, and Self-esteem
67
Krate, R.. Leventhal, G., and Silverstein, B. (1974).
transformation of negro-to-black identity.
, 35, 1071-1075.
Self-perceived
Landreth, C. and Johnson, B. (1953). Young children's responses to
a picture and inset test designed to reveat reactions to
persons of different skin color.
24, 63-79.
.
LaViolette, F. and Silvert. K. (1951). A theory of stereotypes. Social
Forces, 29, 257-262.
Linville, P. and Jones, E. (1980). Polarized Appraisals of Out-group
members.
5, 689-703.
I
Lipprnann, W. (1922). Public Opinion. New York : Harcourt, Brace
Jovanovitch.
Marcia, J. (1980). Identity in adolescence, In J. Adeison,
. New York : John Wiley and Sons.
McAdoo, J. (1971). An exploratory study of racial attitude change
in black preschool children using differential treatments.
Dissertation Abstracts, 31, 3963-A.
McConahay, J. (1 982). Self-interest versus racial attitudes as
correlates of anti-busing attitudes in Louisville: Is i t the
, 44, 692-720.
buses or the blacks?
McConahay, J. (1986). Modern racism, ambivalence, and the Modern
Racism Scale. In Dovidio, J. and Gaertner, S. (Eds.),
Discrimination and Racism, Orlando: Academic Press.
McDonald' K.A. (1980). Enhancing a child's positive self-concept. In
. Provo :
T.D. Yawkey (Ed.),
Brigharn Young University Press.
Milner. D. (1975).
Publications.
I
,
,
( I ) .
.
Beverly Hills : Sage
Racial prejudice. In J. Turner and H. Giles,
. Chicago : The University of Chicago
Press.
Morland, J. (1958). Racial recognition by nursery school children in
Stereotypes, Beliefs, and Self-esteem
68
Lynchburg, Virginia.
37, 132-137.
Parham, T. and Helms, J. (1985). Attitudes of racial identity and
self-esteem of black students : An exploratory investigation.
, 26, 143-147.
Paul, M. and Fischer, J. (1980). Correlates of self-concept among
black early adolescents.
,
9, 163-173.
Peterson, C. and Austin, J. (1985). Review of Coopersmith selfesteem inventories. In J. Mithchell, Jr. (Ed.), 9th M e w
Measurements Yearbook, 1, 396-397.
Phillips, D. and Zigler, E. (1980). Self-concept theory and its
practical implications. In T. D. Yawkey (Ed.), "The Self-Conce~t
~f the Youna Child. Provo : Brigharn Young University Press.
Phinney, J. (1989). Stages of ethnic identity development in
minority group adolescents.
34-49.
, 9,
Phinney, J. (1990). Ethnic identity in adolescents and adults :
Review of research. Psvcholoaical Bulletin, 108, 499-51 4.
Phinney, J. and Alipuria, L. (1990). Ethnic identity in college
students from four ethnic groups. J
171-1 83.
Porter, J. (1971). Black child, white child.
University Press.
Cambridge : Harvard
Proshansky, H. (1966). The development of intergroup attitudes.
1.W. Hoffman and M.L. Hoffman (Eds.), &view of child
jjevelopment research V O ~(2). New YO& : Russell Sage
Foundation.
In
Pushkin, . (1967). A study of ethnic choice in the play of young
,
children in three London districts.
University of London, London, England. (Cited in P. Katz, 1976).
Radke-Yarrow, M., Trager, H., and Miller, J. (1952). The role of
parents in the development of children's ethnic attitudes.
Child Develoament 23, 13-53.
Stereotypes, Beliefs, and Self-esteem
69
Rosenberg, M. and Simmons, R. (1971).
Washiqgton, D.C. : American Sociological Association.
Samuels, S. (1977).
New York : Human Sciences Press.
Samuels, D. and Griffore, R. (1979). Ethnic and sex differences In
self-esteem in preschool children.
Psycholo~l~
135,
, 33-36.
Secord, P. (1 959). Stereotyping and favorableness in %theperception
of Negro faces.
59
309-315.
I
Sherif, M. and Cantril, ti. (1947).
involvements. New York : John Witey and Sons.
Smith, E. (1989).
concerns.
Black racial identity development: Issues and
, 17(2), 277-288.
Spatz, K., and Johnson, F. (1973). Internal consistency of the
Coopersmith Self-Esteem Inventory.
Psvc holoaical Measurement, 33, 875-876.
Stevenson, H. and Stevenson, N. (1960). Social interaction in an
interracial nursery school.
61, 37-75.
Stevenson, H. and Stewart, E. (1958). A developmental study of
racial awareness in young children.
, 29,
399-409.
Tajfel, H. (1 969). Cognitive aspects of prejudice.
Issues, 25(4), 79-97.
Tajfel, H. and Turner, J. (1986). The social identity theory of
intergroup behavior. In S. Worchel, and W. Austin (Eds.),
. Chicago : Nelson-Hall
Publisher.
Trager. H. and Yarrow, M. (1952). They learn what thev livg. New
York : Harper and Brothers Publisher.
Vaughan,
G. (1964). Ethnic awareness in relation to minority group
I
Stereotypes, Befieis, end Self-esteem
70
membership.
, 105, 119-130.
Vaughan, G. (1986). Social change and racia! identity: Issues in the
use of picture and doll measures.
P s v s h o l o p ~ ,38(3), 359-370.
Vinacke, W. (1957). Stereotypes as social concepts.
Social Psvcho\~av,46, 229-243.
White, C. and Burke, P. (1887). Ethnic role identity among black and
white college students : An interactionist approach.
Socioloaical P e r s ~ e c t i v e ~30,
, 310-331.
Wyne, M., White, K., and Coop, R. (1974). The Black Self. New Jersey:
Prentice-Hall, Inc.
Yeatts, P. P. (1968). An analysis of developmental changes in the
self-report of negro and white children grade 3-12.
Abstracts, 29, 823-A.
t nsltucl r ans :
staterrient descrl9er: how you icsho!iy feel, put an :: b the
column 'LIk He-' lf a sta?emecit dscs nat descrlbc hcw you
usually feel, put an X'fn the calilrni~'Unl lZe Me.. There are BG
rlglri ui- wrong answers. Oegin a t qncs&lon number 1 wd mar!:
)f a
a1 S 25 statements.
Like
flc
t-l
4
F
-
; can make l.ip inv mlr,C wlthOUi
15. I'm 3
10: 3:
tufi t O Of-
tor! much tr*~tible
~ilth
6. I OPf !-!gSer egz; iy 2 t ROMe
; 7 !r [a::l;'c:r:;? 2 ign; ::;w? t o 3?[ usc.5 t o anythir;g fits.%
Questionnaire:
Check the point on the graph which applies for you.
Your answers w i l l not be connected with
your identity, therefore you may provide honest replies.
I t is also okay if you have
arcvlousIy completed this questionnaire.
1. Over the past few years. the government and news media have shown more respect to blacks than they
deserve.
I
1
Strongly
Disagree
Moderately
Disagree
-
I
Neither Agree
nor Disagree
Moderately
Agree
-
I
I
Suongly
Agree
It is easy to understand the anger of biack people in America.
2.
I
Strongly
Disagree
I
Moderately
Disagree
I
Neither Agree
nor Disagree
I
Moderately
Agree
I
Strongly
Agree
3. Discrimination against blacks is no longer a problem in the United States.
I
I
Slrongly
Disagree
Moderately
Disagree
I
Neither Agree
nor Disagree
I
Moderately
I
Strongly
Agree
Over the past few years, blacks have gotten more economically than they desetve.
4.
I
Strongly
Disagree
5.
I
Moderately
Disagree
I
Neither Agree
nor Disagree
-
I
Moderately
Agree
I
Strongly
Agree
Blacks have more influence upon school deskgregation plans than h e y ought to have.
I
I
Strongly
Disagree
Moderately
Disagree
I
Neither Agree
nor Disagree
I
Moderately
Agree
I
Strongly
Agree
6 . Blacks are getting too demanding in their push for equal rights.
I
Strongly
Disagree
I
Moderate! y
Disagree
I
Neither Agree
nor Disagree
I
Moderately
Agree
1
Strongly
Agree
Blacks should not push themselves where they are not wanted.
7.
I
Strongly
Disagree
I
Moderarely
Disagree
I
Neither Agree
nor Disagree
I
Moderately
Agree
I
Strongly
Agree
This audp is inqicstip~tirlg
hcw pe~pleperceive the re!ziioncns between words.
il' you &side lo pwtJcipale. you wfil be oat?d about !30 stuck~td
pYlicipsrhg in this srvdy. For iPie nng, about ul how aS your t h e will 3e
nzedcd.
Iln rhi:: study, ydu mil see a series of words appmr on e cornpuler sneee.
You wii: be a k e d to mpwd yes t;r. no to wkether tb.2 mmnd word can elrer
be 4~~ of the first word.
>
-
You may withdraw from t h i s stucd'j at my time witsout peamy or prejudik.
Your p~ilrjcipatiooor i& of gkartidpatioa wi!i in uo way aMecl your grade in
Introduclcn- P c h o I o a - . 18 yola t i ~ zng
~ e qumtions thst can not be
mstpered by rhe +?:{perineater.ptezze fw?free ro cvjrriast &. Juz3 S'Uen.
2'7 1 -286 2. Obln 3 i 7,Dr&e U~iv~rsity.
f-
'
-
74
INSTRUCTIONS
This is a computer task in which you will be asked to make yes or no
responses.
You will see adjectives that describe people and adjectives
that describe houses.
Each adjective will be followed by one of the three
words:
Black, White, or Me. You are asked to respond yes or no to whether the
adjective could ever be descriptive of the word that follows.
I will have you first go through some practice trials in order to
become familiar with the procedure.
Are there any questions before we begin?